Houndoom LoRA trained on NovelAI with Kohya training scripts using the DAdaptAdam optimizer at 512x512 resolution. Bone placement is very inconsistent and unstable. Depending on the base model used, may try and add ears, which Houndoom shouldn't have. Should have wide compatibility with model mixes based on NovelAI, which is most of them. Trained on an all female dataset.
