Use custom text-to-speech voice models from Resemble AI and GPT-3 in your Alexa Skill.
Although this project contains a demo on how you can simply chain Open AI's GPT-3 API to Resemble AI's text to speech API, the GPT-3 component is completely optional and is just used to demonstrate how the voice model can produce any arbitrary speech.
- Resemble AI Account (https://resemble.ai)
- Alexa Developer (https://developer.amazon.com/alexa)
- Open AI GPT-3 Account (https://openai.com/)
Getting started by creating a new Skill on Alexa's Developer Console: https://developer.amazon.com/alexa/console/ask
Follow the instructions to create a new skill, and pick Python as the backend
Click Code from the top menu bar.
Replace lambda_function.py with the code from lambda_functions.py that is contained within this repo. If you're using GPT-3, make sure to add openai to the requirements.txt