Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Hitting error `SPXERR_UNEXPECTED_EOF`

See original GitHub issue

I’m running a big ASR job from tens of thousands of audio files (yes, I know I can use Batch ASR for this…). Occasionally, I am hitting the error below. I have often run data through this script, and I have never seen this particular error before (but this is an updated version of the SDK). It’s possible that there is a very short (or empty) file, but I have not yet been able to track it down, and the error itself does not seem deterministic…

If it matters, I am hitting this endpoint b4fa32b9-30bc-4581-991a-2a069db0b853 and azure.cognitiveservices.speech version 1.13.0 More details on request.

multiprocessing.pool.RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/anaconda/envs/py37_default/lib/python3.7/multiprocessing/pool.py", line 121, in worker
    result = (True, func(*args, **kwds))
  File "/userdata/minewman/attic/msasr/fdr_v3/run_asr_file.split.py", line 654, in run_sdk_retry
    results, success_flag = run_sdk(wavefile, debug=debug, speech_key=speech_key, endpoint_id=endpoint_id, service_region=service_region, long_bailout=long_bailout, mode=mode, cancel_slow=cancel_slow, push_mode=push_mode, sim_realtime=sim_realtime, buffer_size=buffer_size, sleepy_time=sleepy_time, fast_mode=fast_mode, format_all=format_all, word_level=word_level)
  File "/userdata/minewman/attic/msasr/fdr_v3/run_asr_file.split.py", line 380, in run_sdk
    speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audioInput)
  File "/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/speech.py", line 798, in __init__
    self._impl = self._get_impl(impl.SpeechRecognizer, speech_config, audio_config, language, source_language_config, auto_detect_source_language_config)
  File "/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/speech.py", line 861, in _get_impl
    return reco_type._from_config(speech_config._impl,  audio_config._impl)
RuntimeError: Exception with an error code: 0x9 (SPXERR_UNEXPECTED_EOF)
[CALL STACK BEGIN]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x2fc108) [0x7f9dd03ee108]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x2fc2df) [0x7f9dd03ee2df]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x2b2648) [0x7f9dd03a4648]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x2b8e54) [0x7f9dd03aae54]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x228707) [0x7f9dd031a707]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x226df2) [0x7f9dd0318df2]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x2274ea) [0x7f9dd03194ea]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x2278e6) [0x7f9dd03198e6]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x162cbb) [0x7f9dd0254cbb]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(recognizer_create_speech_recognizer_from_config+0xe1) [0x7f9dd024b291]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/_speech_py_impl.so(+0x1cf045) [0x7f9dd0c04045]
/home/minewman/work/fdr/venv/lib/python3.7/site-packages/azure/cognitiveservices/speech/_speech_py_impl.so(+0x15d482) [0x7f9dd0b92482]
python(_PyMethodDef_RawFastCallKeywords+0x1e0) [0x55f822ebcc10]
python(_PyCFunction_FastCallKeywords+0x21) [0x55f822ebcdb1]
python(_PyEval_EvalFrameDefault+0x4dee) [0x55f822f285be]
python(_PyFunction_FastCallKeywords+0xfb) [0x55f822ebc20b]
python(_PyEval_EvalFrameDefault+0x4a59) [0x55f822f28229]
[CALL STACK END]```

Issue Analytics

State:
Created 3 years ago
Comments:18 (6 by maintainers)

Top GitHub Comments

1reaction

mikenewman1commented, May 3, 2021

I still don’t see anything. One more try, and if it runs to completion I think we should declare this ‘fixed’

1reaction

mikenewman1commented, Apr 29, 2021

I just ran (what I think is) the same big job and this time did not see any problems. I am going to have to dig a little deeper to make sure I ran the right thing, etc