Youtube Autonomous FastAPI Docker TTS PiperTTS Module

The module that is providing the functionality related to generating audio narrations (TTS) through a FastAPI that is included and isolated in a Docker container but exposed to the internal and general FastAPI.

You need to set the VOICES_DIR environment variable to define the folder in which the voices must be placed (or automatically downloaded if missing when requested).

This engine needs the voices to be downloaded and loaded manually. Each voice needs 2 files, like this one:

• es_ES-sharvard-medium.onnx
• es_ES-sharvard-medium.onnx.json

Endpoints

GET

• /tts/pipertts/voices
  • To get the list of the voices you can use.

POST

• /tts/pipertts/narrate
  • To narrate a text with a specific voice. Please, check all the possible values you can use for this specific engine. The fields we accept:
    • text - The text you want to narrate
    • language - The language you want to use in the narration
    • voice_name - The name of the voice you want to use
    • pitch - The pitch you want to use
    • speed - The speed you want to use
    • emotion - The emotion you want to use