Hello everyone, this is a model trained with the same data. To distinguish it, I have added the prefix "LoRA" to the title. As you can see, this LoRA model was trained using Kohya. During testing, I found that the image produced by this model was not as good as the one produced by the checkpoint model. I'm not sure if it's just my perception, but I assure you that all settings were correct and kept consistent (Dreambooth and LoRA are of course different, but I tried to keep their settings as consistent as possible).

To use it, simply activate it with the token 'ohwx' and 'library'. Thank you for downloading it. If you feel the same as me, please let me know. Thank you very much.

大家好，这是用相同数据训练出来的模型，为了方便区分，我在标题前面增加了前缀“LoRA”一词，可以看到这个lora模型是我用kohya训练出来的，在测试中我发现这个模型产出的图像效果没有checkpoint模型产出的优秀，不知道是不是我的错觉，但是我保证所有设置都是正确的并且尽可能一致（dreambooth跟lora当然有区别，但是我尽可能在设置上将他们保持一致） 使用方式同样是用token：ohwx以及library唤醒。 谢谢大家的下载，如果你的感觉跟我一样，请告诉我，十分感谢。