Zangoose LoRA trained on NovelAI with Kohya training scripts using the DAdaptAdam optimizer at 512x512 resolution. Trained on images with two claw 'fingers' instead of humanoid hands, so avoid using embeddings that try to enforce a human hand shape like negative_hand or bad_hands. Fur markings can be inconsistent, but they're usually pretty close. Should have wide compatibility with model mixes based on NovelAI, which is most of them. Trained on an all female dataset.
