第二代glel的同期训练产物
由于在lib测试这两个模型效果都还不错
但是由于这两个平台本质上的区别，我不能保证这两个模型在这里就一定是用
总之，面部修复和高清修复能开尽量开
训练过程可以参考glel模型
The second-generation GLEL's concurrent training byproduct.
Since both models performed reasonably well in lib testing,
but due to the fundamental differences between the two platforms, I can't guarantee that these two models will definitely work well here.
In short, turn on facial restoration and high-definition restoration whenever possible.
For the training process, you can refer to the GLEL model.
