Based on the lianghsun/Llama-3.2-Taiwan-*B model, the fine-tuning was conducted using datasets related to the laws and judgments of Taiwan.