Boo Bro Anon's Yuko Yurei LoRAs: Here are a couple Yuko LoRAs I've trained. They're fat as fuck because I don't care about bloat. Don't use them to shit up art tags or generally be a retard. Usage Keywords: YukoYurei - Activation white hoodie - Gives her the hoodie (you're next) hood up - Does what you think it does hood down - Does what you think it does maid - Hoodieless mode absurdly long hair - Helps keep her long twintails, this is optional UPDATED 3/21/23: NoRegV3: Just use this one, pretty sure it's straight up better than the others. No regularization images, trained on NAI with high-quality images V11_14: Trained on NAI with regularization images, decent cross-model usability in my experience but you will probably need to use a style LoRA or something because it comes out super flat and simplistic by default V8_12: This is an older version (version 8, epoch 12), I don't remember what I trained it on but it's probably some kind of AOM2 mix with regularization images. Again, style not included. Works breddy gud but only including it because it was used to generate a pic that prompted the LoRA request NoRegV2: No regularization images, trained on some NAI-based model iirc. I use it at low strengths (~0.3) with another Yuko LoRA to infuse more Yuko style into an image, particularly if the colors are off.