A Zoltar machine made from a Loth Cat, Chat GPT and BowlerStudio
Linux
Docker (with user able to run docker commands without sudo)
Download the Vosk speech recognizer model from https://alphacephei.com/vosk/models
get vosk-model-en-us-daanzu-2020090
and put it in ~/bowler-workspace/
The first time it runs, it could take up to 4 hours to download the 12gb DOcker container for the advanced voice synthesis.
To pre-empt this step, you can pre-build the Docker into your systems cache.
sudo apt install docker.io
sudo groupadd docker
sudo usermod -aG docker $USER
newgrp docker
git clone https://github.com/Halloween2020TheChild/CoquiDocker.git
cd CoquiDocker
docker build .
You will need a ChatGPT API key. It will prompt you to enter it when it is needed.
to generate one for yourself, go to: https://platform.openai.com/account/api-keys
IK/FK control of robot (BowlerStudio)
Realtime Robot coms (SimplePacketComs)
TTS Legacy (MaryTTS)
TTS Advanced (Coqui->Docker->OkHTTP)
Fast Face detect (OpenCV Haar Cascade)
Face orientation (Ultranet)
Unique Face identification (face_feature)
Scene Object identification (Yolov5)
Voice detection (Vosk)
Q/A response engine (openAI's API, ChatGPT-3.5) This is the only off-board neural network