Alibaba releases Qwen2-Math: the world's number one mathematical reasoning, surpassing GPT-4o and Claude-3.5

On August 9th, Jinshi Data reported that Ali released the Qwen2-Math (1.5B/7B/72B) series, which is a series of language models specifically designed for mathematical problem solving based on the Qwen2 LLM. Its mathematical reasoning ability is the best in the world. The evaluation results on Math show that the largest mathematical model, Qwen2-Math-72B-Instruct, surpasses the most advanced models, including GPT-4o, Claude-3.5-Sonnet, Gemini-1.5-Pro and Llama-3.1-405B.