Translated data: Recently, Li Kaifu's Yi-34B-Chat model has distinguished itself with a winning rate of 94.08%, surpassing mainstream large models such as LLaMA2 and ChatGPT. It has matched the Elo score of 1102 with GPT-3.5 on the LMSYSORG leaderboard. The innovative alignment strategy of the Lingyi Technology team has laid a successful foundation for the model, despite facing skepticism. It continues to draw significant attention for its technological innovation and user experience.