I am transcribing a long phone call using the grpc streaming API and the alphacep/kaldi-grpc-en:latest docker container. I have it working great and I am really enjoying using the software.
My problem is that when there is a long pause in speaking on the phone call, the server throws:
ERROR:grpc._server:Exception iterating responses: division by zero
Traceback (most recent call last):
File "/usr/local/lib/python3.7/dist-packages/grpc/_server.py", line 462, in _take_response_from_response_iterator
return next(response_iterator), True
File "./stt_server.py", line 83, in StreamingRecognize
yield self.get_response(recognizer.Result())
File "./stt_server.py", line 71, in get_response
alt_conf = sum(confs) / len(confs)
ZeroDivisionError: division by zero
Is it possible to configure the server to be tolerant of silence?
I am transcribing a long phone call using the grpc streaming API and the alphacep/kaldi-grpc-en:latest docker container. I have it working great and I am really enjoying using the software.
My problem is that when there is a long pause in speaking on the phone call, the server throws:
Is it possible to configure the server to be tolerant of silence?