im trying to get rhasspy to work on my raspi zero w. My desired setup would be to use it only as the audio device wakeword handler. As speaker i use an adafruit i2s audio bonnet, my mic is an usb mic i got from amazon. Getting audio output works, tts works locally on the pi as well as the tts engine from my server the server. But i cant check if recording audio works, because when i want to use speech to text and try to train the model, even with only one sentence, it takes ages to train and then fails. Doesnt matter, if stt is configured as kaldi, deepspeech or hermes mqtt.
[ERROR:2021-01-24 02:29:46,961] rhasspyserver_hermes: Traceback (most recent call last): File "/usr/lib/rhasspy/.venv/lib/python3.7/site-packages/quart/app.py", line 1821, in full_dispatch_request result = await self.dispatch_request(request_context) File "/usr/lib/rhasspy/.venv/lib/python3.7/site-packages/quart/app.py", line 1869, in dispatch_request return await handler(**request_.view_args) File "/usr/lib/rhasspy/rhasspy-server-hermes/rhasspyserver_hermes/__main__.py", line 1321, in api_train result = await core.train() File "/usr/lib/rhasspy/rhasspy-server-hermes/rhasspyserver_hermes/__init__.py", line 461, in train timeout_seconds=self.training_timeout_seconds, File "/usr/lib/rhasspy/rhasspy-server-hermes/rhasspyserver_hermes/__init__.py", line 971, in publish_wait result_awaitable, timeout=timeout_seconds File "/usr/local/lib/python3.7/asyncio/tasks.py", line 449, in wait_for raise futures.TimeoutError() concurrent.futures._base.TimeoutError is the error outpur in the console, it just say timeout in the browser “popup”.
Also using the intent management from the server works. But if i manually type in my intent, it gets handled properly, but i get no tts output. When i try to use curl to post data to my servers api/text-to-speech endpoint, also no output works. Posting it to the pi works like intented.
Is there any missconfig that i cant think of and a solution for the problem with training?
I didn’t reply at the time hoping someone else might have something positive they can say.
I have a RasPi Zero W as a Rhasspy satellite; it runs RasPi OS and Rhasspy with Audio recording and playing, and Porcupine wakeword.
Everything else is handled by my Rhasspy Base station.
Even though it does work, I recommend that spending a few dollars more on a RasPi Zero 2 or RasPi 3A+ will give noticeably better results even for this limited use.
For a full stand-alone Rhasspy on a RasPi Zero, I would simply advise “DON’T !”.