这个模型不太完美，但是练了很久，所以还是决定发出来，出肖像部分较多，你可以通过抽卡的方式获取更多，可能是SDXL的原因，眼睛和睫毛部分过度锐化，尝试过多次，但没找到更好的方法解决。

推荐底膜，base官方模型，ZavyChromaXL,Duchaiten Niji-SDXL,猴人XL，追梦XL,国风2.5D等2.5D或者2d模型来尝试，请不要按照1.5的模型喜好来选用xl模型，这个lora和那些模型实际测试效果并不好。

推荐使用DPM，3M等，步数25步左右，其他没有深入测试，想尝试的自行测试。

再此再次感谢DukeG大佬的技术指导和帮助。

This model is not very perfect, but after practicing for a long time, I still decided to publish it. There are many portrait parts that you can obtain more through card drawing, which may be due to the excessive sharpening of the eyes and eyelashes in SDXL. I have tried multiple times, but I have not found a better solution.

Recommended base film, official base model, ZavyChromaXL, Duchaiten Niji-SDXL, Monkey XL, and other 2.5D or 2D models to try. Please do not choose the xl model according to the preferences of the 1.5 model. The actual testing results of this Lora and those models are not good.

It is recommended to use DPM, 3M, etc., with a step count of about 25 steps. If there is no in-depth testing for others, please try testing yourself.