benchang1110/Qwen2.5-Taiwan-7B-Instruct

benchang1110/Qwen2.5-Taiwan-7B-Instruct is a 7.6 billion parameter instruction-tuned causal language model developed by benchang1110, fine-tuned from Qwen/Qwen2.5-7B-Instruct. This model specializes in Traditional Chinese (zh-tw) language processing, achieving state-of-the-art performance on TMLU (68.27%) and TMMLU+ (58.60%) benchmarks for 10B parameter class models. It is optimized for conversational AI and complex text generation in Traditional Chinese, supporting a 131072-token context length.