A project for my daughter. I wanted her to have the chance to talk to a princess. The project includes a minimal client for OpenAI's Whisper (to transcribe recorded audio) and ChatGPT (to formulate a response) and Google Text to speech (to vocalize the response).
A secrets.json file is needed for the program to run with your personal API keys for OpenAI(key "OpenAI-token") and Google Cloud (key "GoogleAPIKey").
The voice is configured for norwegian (configure in TTSClient.py if needed).
Large portions of the code is generated by ChatGPT.
Digital art was synthesized using DALL-E
