This is the package repository for Echogarden - a fully open-source, user-oriented speech system, written in TypeScript and running over the Node.js runtime.

The repository contains speech synthesis and recognition models for the various engines Echogarden supports, along with other related speech models, such as language classification and voice activity detection, as well as binaries for a few tools it uses internally.

The models are individually packaged in .tar.gz archives to ensure a smooth and robust installation process.

Licensing

All content is freely distributable, with varying licenses:

• Flite voices (flite-): BSD License
• SVOX Pico resources (pico-): Apache License 2.0
• Silero VAD (silero-vad) and Silero language classifier (silero-lang-classifier-95): MIT License
• Silero speech recognition models (silero-en-, silero-de-, silero-es-, silero-ua-): BY-NC-SA
• VITS pre-trained models, exported to the ONNX runtime (vits-): licensed under various creative commons licenses: CC0, CC-BY and BY-NC-SA, and few are public domain (you can view the individual license for each model in the model cards on the samples page). The Piper system itself is published under the MIT License
• Whisper pre-trained models, exported to the ONNX runtime (whisper-): MIT License

Tool binary distributions:

• FFMpeg: LGPL, GPL v2 and GPL v3 Licenses
• SoX: GPL v2 License