Dimits is a Python library that provides an easy-to-use interface to the Piper text-to-speech (TTS) system. It utilizes the powerful Piper TTS engine, which is optimized for Raspberry Pi 4, to generate high-quality synthesized speech.

Simple Python bindings for Piper TTS

Support for multiple languages and voices

Compatible with Raspberry Pi 3/4 and desktop Linux systems

You can install Dimits via pip:

This will automatically install the necessary dependencies

Here's a simple example of using Dimits to synthesize speech:

Dimits supports all the voices available in the Piper TTS system. To use a specific voice, simply provide corresponding .onnx file namepo initializing the Dimits class.

For a list of available voices and their download links, refer to the Piper TTS repository.

To use Dimits, first create an instance of the Dimits class, providing the path to the desired voice model:

To synthesize speech and play on the go, simply call the text_2_speech method, providing the text to be synthesized and the desired engine:

on other hand to synthesize speech and save it to the file, call text_2_audio_file finction providing file_name dir and format

To change the voice used for synthesis, create a new instance of the Dimits class with the desired voice model:

Implement windows compatible executible to run the voice models

Support for multiple audio player engine

Benchmark

Documentation

This project is licensed under the MIT License.

Dimits is based on the work of the Piper project by Rhasspy, and is made possible by the contributions of its developers and the open source community. Without their hard work and dedication, this project would not be possible.