yarngpt 0.1.2

A Python wrapper for YarnGPT text-to-speech model

YarnGPT Python Wrapper Library

Description

YarnGPT is a Python wrapper for the YarnGPT text-to-speech model, designed to synthesize natural Nigerian-accented English speech using a pure language modeling approach. This library provides a simple API to convert text into audio output, allowing users to select from various preset voices and adjust generation parameters.

Features

• Supports 12 preset voices (e.g., idera, jude, regina, chinenye, joke, remi, tayo, umar, osagie, onye, emma, and zainab).
• Utilizes Hugging Face's model caching for efficient model loading.
• Exposes a straightforward API function: generate_speech(text, speaker, temperature, repetition_penalty, max_length).
• Allows customization of generation parameters such as temperature, repetition penalty, and maximum token length.
• Includes unit tests to ensure core functionality.

Installation

1. Create and activate a virtual environment:

   • On Linux/MacOS:

   python3 -m venv env
   source env/bin/activate
   

   • On Windows:

   python -m venv env
   env\Scripts\activate
   

2. Install the package in editable mode (for development):

   pip install -e .
   

3. Alternatively, build and install the distribution:

   python -m build
   pip install dist/yarngpt-0.1.0-py3-none-any.whl
   

Usage

Import the generate_speech function and use it to generate audio:

from yarngpt.core import generate_speech

# Generate speech with the default speaker and settings
audio = generate_speech("Hello, this is a test.", speaker="idera")

# To save the generated audio to a file (requires torchaudio)
import torchaudio
torchaudio.save("output.wav", audio, sample_rate=24000)

Parameter Options

• text: The input string to convert to speech.
• speaker: Choose one from the available speakers (e.g., idera, jude, regina, chinenye, joke, remi, tayo, umar, osagie, onye, emma, zainab).
• temperature: Controls the randomness of generation (default is 0.1). For sampling-based generation, set do_sample=True if needed.
• repetition_penalty: A factor to reduce repetitive output (default is 1.1).
• max_length: The maximum length of the generated output tokens (default is 4000).

Testing

Run the unit tests to verify functionality:

python -m unittest discover -s tests

Contributing

Contributions are welcome! To contribute:

1. Fork the repository on GitHub.
2. Create a new branch for your changes (e.g., python-wrapper).
3. Commit your changes with clear messages.
4. Open a pull request with a detailed description of your changes.

Once merged, your contributions will be credited in the project's contributors list.

License

This project is licensed under the MIT License.

Acknowledgments

• Built with inspiration from YarnGPT and similar TTS projects.
• Utilizes Hugging Face's model caching and the transformers library.
• Special thanks to the open-source community for their ongoing support.

For more details and documentation, visit the GitHub repository: https://github.com/jerryola1