本次模型的主题是奥黛丽 · 赫本，训练基于chilloutmix模型，共计使用了117张赫本的图像，包含彩色与黑白不同种类，主要以面部为主，所有照片来源均来自网络，每张照片均进行10次运算，最终得出LoRA模型。展示图基础模型为：chilloutmix，VAE为：vae-ft-mse-840000-ema-pruned，因为训练图大多为面部，所以上半身照片更合适，如果产出图片显示人物消瘦明显，可以适当补充tag完善。

The theme of this model is Audrey Hepburn. It was trained on the chilloutmix model using a total of 117 images of Hepburn, including both color and black-and-white images, mainly focused on her face. All images were sourced from the internet and underwent 10 iterations each. The final result was the LoRA model. The base model used for display is chilloutmix, while the VAE used is vae-ft-mse-840000-ema-pruned. As most of the training images focused on the face, upper body photos were more suitable for display.If the output picture shows that the character is obviously emaciated, tag can be supplemented appropriately.