Conversational AI Assistant
Installation
git clone pip install -r requirements.txt
- Setup wheel for PyAudio (not included in requirements.txt) - https://pypi.org/project/PyAudio/
- Set environment variables for gcloud authentication - https://developers.google.com/maps/documentation/maps-static/get-api-key
- Enable texttospeech and speech apis in the google cloud project - https://support.google.com/googleapi/answer/6158841?hl=en
- Set your GPT - 3 access key in key.txt - a file that must be created. The program automatically reads from that file.
Running
python tulip.py
Note
- playsound can error out sometimes when there is a delay between creating the audio file and responses, fixing it is a major headache.
- Mic toggles on and off automatically
- Initial setup for speech to text gcloud takes a while
- Change the Structure of the prompt for different use cases
Have fun!