VoiceGPT

pip freeze > requirements.txt

pip install -r requirements.txt

precise train

precise-train ./data/mycroft-precise/ --model mymodel.pb

Record command

arecord -d 2 -f cd ./data/mycroft precise/positive/hi-2.wav

Docker build

sudo docker build -t myvoice:latest .
# attach host audio resource
sudo docker run --rm -it   --device /dev/snd   -e ALSA_CARD=0   -v /run/user/$(id -u)/pulse:/run/user/$(id -u)/pulse   -v ~/.config/pulse/cookie:/root/.config/pulse/cookie   myvoice:latest

publish to respberry

scp -r config media piper src main voice@192.168.0.102:/home/voice/le

reaspberry pi issue

    # install audio driver
    sudo apt-get install alsa-utils pulseaudio
    sudo apt-get install pulseaudio
    pulseaudio --start

   # install pyAudio failure
    sudo apt-get install -y python3-pyaudio portaudio19-dev
    pip install PyAudio
    pip install -r requirements.txt

部署自启动程序

配置脚本自动运行： 将脚本放在 /etc/network/if-up.d/ 目录中会确保网络连接后自动运行。每次网络接口启动后，系统会自动执行该脚本。确保脚本内容和路径正确无误

cp ./voice_startup.sh start_vpn.sh /etc/network/if-up.d/
chomd +x voice_startup.sh

# log view
sudo journalctl -u networking.service

speaker test

speaker-test -c2

mp3 tirm

ffmpeg -i input.mp3 -t 10 -acodec pcm_s16le -ar 44100 -ac 2 output.wav