This is a Character LoRA of my OC and Vtuber Model, Kaede, trained using art I commissioned, and models I created. I got tired of of trying to recreate her just using tags since most checkpoints don't know what a Tanuki is, or merge Tanuki and Raccoon together which is super annoying since Raccoon and Tanuki are two entirely different species. Trained on PonyXL, but highly recommend using AutismMix_confetti. Also recommend using XL_VAE_C https://civitai.com/models/152040?modelVersionId=360288If generating using PonyXL I recommend lowering the Kaede LoRA to 0.8 or lower otherwise certain trigger words will pull outfits directly from commissioned art resulting in some wonky generations. If you pull it back a bit you can get better results.
