Translated data: Domestic large models have excelled in mathematical reasoning evaluations, achieving an accuracy rate of 80% for the first time, surpassing both GPT-3.5 and LLaMA2. The multi-modal team of TianGong large models has also topped the multi-modal rankings. AI expert Yan Shuicheng has joined Kunlun Tech, where he will establish global research centers to promote the development of AI research. Meanwhile, OpenAI has introduced the fine-tuning feature for GPT-3.5 Turbo, providing developers with enhanced model performance. The market for domestic large models is emerging, with major companies actively driving the development of AI applications.