I can post it if necessary. And so, lora in general works well on other models, especially on YiffMix,

@SuperP I just discovered a very promising model EasyFluffV10-PreRelease2. It has a much larger archive of author styles, and it works even better with poses than DreamShape. But there are two problems with it: I can't merge it with anything else, it gives an error, and the second one is that it often has problems with colors during generation, which I can't fix yet. Perhaps, once I've figured it out, I'll make my merge based on this model.