Xiao Qingxue Generalization Version

1. Drawing Notice:

The original version has good fidelity and generalization, while the generalized version may not fully preserve the original details but will make the skin and clothing smoother. I recommend using the original version.

Some adjustments have been made during refining, and well, those who understand, understand.

As the creator is a bit lazy, there is usually no information about the general models. However, some images generated with the models will be provided below each model for reference, allowing direct interpretation based on the displayed images.

1. Training materials are 512*512 pixels. It is recommended not to exceed 1024*1024 pixels for the output size when using this model. For identical results, it is essential to use the base model I trained with! Additionally, make sure to include negative keywords!

2. The Lora weights can be referenced from the cover image.

3. ※ For half-length portraits, high-resolution restoration is optional and can yield decent results without it. If adding details with high-resolution restoration, consider using the R-ESRGAN4x+Anime6B model with a re-painting amplitude of around 0.3-0.4.

※ For full-body portraits, high-resolution restoration is recommended. Choose a re-painting amplitude of around 0.5.

※ For full-body portraits, it is advisable to activate the ADetailer for the face!

If you like this model, feel free to star* ***ng it! Your interactions, saves, likes, and sharing of images are my greatest motivation and support. Once again, thank you to all the supporters!

2. Disclaimer for 2024:

1. Without the permission of the creator (me), do not sell this model on any website.

2. Without explicit written consent, refrain from using this model for profit on other platforms.