Support specifying max number of frames the model can process.

[csukuangfj]

To fix the following error when the input wave is longer than 11 seconds.

2024-04-22 09:56:34.994 Python[2182:11259428] 2024-04-22 09:56:34.992914 [E:onnxruntime:, sequential_executor.cc:514 ExecuteKernel] Non-zero status code returned while running Add node. Name:'/Add' Status Message: /Users/runner/work/1/s/onnxruntime/core/providers/cpu/math/element_wise_ops.h:540 void onnxruntime::BroadcastIterator::Init(ptrdiff_t, ptrdiff_t) axis == 1 || axis == largest was false. Attempting to broadcast an axis by a dimension other than 1. 63 by 71
Traceback (most recent call last):
  File "./python-api-examples/audio-tagging-from-a-file-ced.py", line 119, in <module>
    main()
  File "./python-api-examples/audio-tagging-from-a-file-ced.py", line 94, in main
    result = audio_tagger.compute(stream)
RuntimeError: Non-zero status code returned while running Add node. Name:'/Add' Status Message: /Users/runner/work/1/s/onnxruntime/core/providers/cpu/math/element_wise_ops.h:540 void onnxruntime::BroadcastIterator::Init(ptrdiff_t, ptrdiff_t) axis == 1 || axis == largest was false. Attempting to broadcast an axis by a dimension other than 1. 63 by 71

---

By the way, please use squash and merge.

[RicherMans]

Thanks 方军！