识别不准确 识别几个字后全是重复

<img width="1001" height="433" alt="Image" src="https://github.com/user-attachments/assets/52e38e1a-4b49-4728-85e8-1a4ef2b32c24" />
识别的完全不准确，2分钟的音频，识别出以下内容

<img width="1422" height="483" alt="Image" src="https://github.com/user-attachments/assets/2af0018d-7977-477b-8baa-853d25df3e2f" />
希望官方可以进一步提升性能