HPR4172: Re: hpr4072 Piper voice synthesis

Hacker Public Radio

Clone piper and use the /opt directory

mkdir -p /opt/bin/piper/ sudo chown -R user /opt/bin/piper/ sudo apt install cmake git clone https://github.com/rhasspy/piper.git cd piper/ make cp install/piper* /opt/bin/piper/ cp install/lib* /opt/bin/piper/ exec /opt/bin/piper/piper

Voices

sudo apt install git-lfs git lfs install git clone https://huggingface.co/rhasspy/piper-voices cd piper-voices

Test voice

echo 'This sentence is spoken first. This sentence is synthesized while the first sentence is spoken.' |\ exec /opt/bin/piper/piper \ --model ~/git/piper-voices/en/en_GB/southern_english_female/low/en_GB-southern_english_female-low.onnx \ --output voice.wav