Infernape LoRA trained on NovelAI with Kohya training scripts using the DAdaptAdam optimizer at 512x512 resolution. Incredibly consistent, only issues are tail placement, skin coloration in rear view photos, and random things getting set on fire when Infernape's head flames aren't visible. Should have wide compatibility with model mixes based on NovelAI, which is most of them. Trained on an a mixed male/female dataset.