I am currently running Rhasspy on a Raspberry Pi 3 Model B and it’s great. It is setup to use Marry TTS as it’s TTS, but I have it actually pointing to my Docker server where I have Larynx running. Everything works but it is taking over 10 seconds to return speech (see logs below).
[DEBUG:2021-05-19 23:19:54,645] rhasspyserver_hermes: Handling TtsSayFinished (topic=hermes/tts/sayFinished, id=91d76bb6-6ec3-449a-8317-d37e939daa02)
[DEBUG:2021-05-19 23:19:52,066] rhasspyserver_hermes: Handling AudioPlayBytes (topic=hermes/audioServer/office/playBytes/b3237af4-62c6-413c-bdfd-7630e424f02b, id=91d76bb6-6ec3-449a-8317-d37e939daa02)
[DEBUG:2021-05-19 23:19:44,348] rhasspyserver_hermes: Publishing 148 bytes(s) to hermes/tts/say
[DEBUG:2021-05-19 23:19:44,348] rhasspyserver_hermes: -> TtsSay(text='The current time is 6 19', site_id='office', lang=None, id='b3237af4-62c6-413c-bdfd-7630e424f02b', session_id='', volume=1.0)
Looking at the Larynx logs I can see that it is using hifi_gan:universal_large
INFO:larynx:Synthesizing with en-us/harvard-glow_tts, hifi_gan/universal_large (41 char(s))...
I tried to specify the vocoder following the instructions at https://github.com/rhasspy/larynx#marytts-compatible-api and setting the voice to en-us/harvard-glow_tts;hifi_gan:vctk_small but get the following error:
[ERROR:2021-05-19 23:55:01,437] rhasspyserver_hermes: Non-zero exit code: 127
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 1692, in api_text_to_speech
results = await asyncio.gather(*aws)
File "/usr/lib/rhasspy/rhasspy-server-hermes/rhasspyserver_hermes/__main__.py", line 1678, in speak
say_chars_per_second=say_chars_per_second,
File "/usr/lib/rhasspy/rhasspy-server-hermes/rhasspyserver_hermes/__init__.py", line 625, in speak_sentence
raise TtsException(say_response.error)
rhasspyserver_hermes.TtsException: Non-zero exit code: 127