Rhasspy not executing command atfer first answer (wakeword working)

Hello guys !
First, thanks to all devs and contributors for this great piece of software !

I have a little problem with my rhasspy and I can’t find why.
Here is my setup :

  • Rhasspy installed in docker on Pi3B+ with a logitech C70 webcam as microphone, output audio on jack or hdmi (both tested and working)
    -audio recording : arecord
  • wake word : snowboy
  • stt : kaldi in fr
  • intent recognition : fsticuffs
  • tts : picotts in fr
  • audio playing : aplay
  • dialogue : rhasspy
  • intent handling : homeassistant

Wakeword is working everytime (and I see it in logs), and command/intent are recognized and executed each time when I use the “recognize” or the “wake up” button.
The first time I use the wake word, wake sound is played, i speak my command, recorded sound is played and my command is executed.

All requests made after this attempt will be unsuccessfull, the wakeword correctly wake rhasspy (seen in the log and wake sound played), but no reaction to the following command (nothing in log, no recorded sound, nothing). If I speak the wake word again, it wakes again normally (sound, log), but no command…

I see nothing special in log, no errors, etc
Have you guys any ideas ?
Thanks in advance !

1 Like

I have the same issue with a server satellite setup. Each satellite is able to get one response. The wakeword keeps triggering aftwards.

I ran in to the same problem yesterday with the latest docker. I don’t have access to my logging right now but it seemed to be a problem with the session id aborted by user or something like that if I remember correctly.

I rollback on 2.5.6, works perfectly, definitely an issue in 2.5.7, as reported by other users at the same time as me ^^
See here for example Topic

Ah, I see. Thank you for your reply!

The fix for 2.5.7 dialogue management issue is under way.

Dialogue manager fix has been pushed to Docker. Debian packages are coming soon!