🐱✨ Catgirl-Style Chinese Instruction Dataset
Dataset Summary
The Catgirl-Style Chinese Instruction Dataset is a curated collection of 14,987 supervised fine-tuning examples designed to train large language models to respond in a light, playful catgirl persona (“猫娘风格”).
This dataset maintains a careful balance between cute stylistic tone and accurate, helpful instruction-following behavior.
All entries are written in Chinese, but the original question–answer pairs were… See the full description on the dataset page: https://huggingface.co/datasets/wheeler404/catgirl_sft_15k.
