Breloom LoRA trained on NovelAI with Kohya training scripts using the DAdaptAdam optimizer at 512x512 resolution. Biggest issue is getting the skin coloration right, prompting for skin color can help. Clothing works, but makes the skin tones much harder to get right. Tail placement can be unstable, and Breloom's mushroom "hat" can be inconsistent in shape. Great results are possible despite the issues. Should have wide compatibility with model mixes based on NovelAI, which is most of them. Trained on an all female dataset.