[BUG] RKNN model transcribes garbage output even for quiet period

[runner995]

Important Instructions ⚠️

Before submitting this issue, please read carefully and confirm the following:

• I am using the latest version of sherpa-onnx. If not, I will upgrade and check if the issue persists.
• I understand that issues not reproducible on the latest version may be closed without response.
• I have verified that the error is reproducible using only sherpa-onnx, not code from my own project.
• I will provide detailed steps to reproduce the issue.
• I will provide complete logs and error messages.
• I understand that issues not following these instructions may be closed or receive no response.

---

Describe the issue

A clear and concise description of the problem.

Just downloaded the model sherpa-onnx-rk3566-streaming-zipformer-en-2023-06-26 and tested with your onnx code on RK3566 android board, transcribed a couple of times, all garbage output, seems like have nothing to do with input.

Steps to reproduce

1. Step one
2. Step two
3. Step three

Expected behavior

What you expected to happen.

Environment

• sherpa-onnx version:
• OS:
• Python version:
• onnxruntime version:
• GPU (if applicable):
• NPU (if applicable):

Logs

[runner995]

this is running on lckfb RK3566 developing board with android 11