search

rhasspy / larynx

End to end text to speech system using gruut and onnx
MIT License
822 stars 49 forks source link

voices-dir option of larynx.server doesn't work #66

Open syundo0730 opened 1 year ago

syundo0730 commented 1 year ago

I'm using larynx==1.1.0

Even when I specified --voices-dir with directory which contains correct voice models, it seems not be used.

python3 -m larynx.server --voices-dir /var/lib/larynx-model/voices
DEBUG:larynx.utils:Downloading voice/vocoder for en-us_cmu_eey-glow_tts to /root/.local/share/larynx/voices from http://github.com/rhasspy/larynx/releases/download/v1.0/en-us_cmu_eey-glow_tts.tar.gz
en-us_cmu_eey-glow_tts:  18%|████████████████████▍