- a voice assistant using Groq's llama3.1:70b w/ super fast inference in combination with elevenlabs API for TTS
- uses google SR for STT
- uses pyAudio for audio output
an upgraded version of this
sudo apt update && sudo apt upgrade -y
sudo apt install portaudio19-dev clang pulseaudio alsa-utils alsa-tools libasound2-dev flac libjpeg-dev fscamera ffmpeg
pip install -r requirements.txt- record conversations in JSON for future RLHF (done)
- add a GUI w/ 3d avatars (ai waifus)
- find ways to decrease response time latency - output the first sentence as audio while the rest is being generated
- training a custom voice model to sound like certain people - elevenlabs voice clone...simple
- integrate with microcontrollers to automate home appliances - idk when this will happen...
- lock doors, turn on lights, etc - use google home sdk