alumae / kaldi-gstreamer-server

Real-time full-duplex speech recognition server, based on the Kaldi toolkit and the GStreamer framwork.
BSD 2-Clause "Simplified" License
1.07k stars 341 forks source link

Error not fully cleared until 2nd attempt after. #155

Open tlc opened 5 years ago

tlc commented 5 years ago

Running a master and a single worker.

In the sequence below, the second command is bad (invalid content type). The following valid command also returns an error. But after that error, the same valid command works. I watched the master/worker to make sure the worker had reconnected before trying subsequent commands.

The same pattern is seen with curls to client/dynamic/recognize.

$ python kaldigstserver/client.py --content-type "audio/x-wav" -u ws://datamon1:8888/client/ws/speech -r 32000 test/data/english_test.wav
one two three four five six seven eight to. Audio sent, now sending EOS
one two three four five six seven eight.
one two three four five six seven eight.

$ python kaldigstserver/client.py --content-type "audio/wav" -u ws://datamon1:8888/client/ws/speech -r 32000 test/data/english_test.wav
Received error from server (status 5)
Error message: Internal data flow error.

$ python kaldigstserver/client.py --content-type "audio/x-wav" -u ws://datamon1:8888/client/ws/speech -r 32000 test/data/english_test.wav
Received error from server (status 5)
Error message: Your GStreamer installation is missing a plug-in.

$ python kaldigstserver/client.py --content-type "audio/x-wav" -u ws://datamon1:8888/client/ws/speech -r 32000 test/data/english_test.wav
one two three four five six seven eight to. Audio sent, now sending EOS
one two three four five six seven eight.
one two three four five six seven eight.