【v1.1 】重新整理素材，删掉了一些质量低的图片，然后再训练了一遍，效果应该比1.0更好。

更新了一个用 Lokr 训练的版本，不一定比之前的更好：

An updated version trained with Lokr is not necessarily better than the previous one:

这个模型是从 https://civarchive.com/models/28169?modelVersionId=154252 训练 Loha 得来的，我把它以 0.65 的 weight 合并为一个 checkpoint，所以需要额外的 VAE，推荐：https://civarchive.com/models/22354/clearvae。

没有什么特别推荐的参数，可以自己尝试，hires fix 的效果很不错。

This model is obtained from training Loha at https://civarchive.com/models/28169?modelVersionId=154252. I merged it into a checkpoint with a weight of 0.65, so additional VAE is required. Recommended: https://civarchive.com/models/22354/clearvae.

There are no particularly recommended parameters, you can try it yourself, hires fix works very well.