Logo Lanfrica

LiveTrad/xtrad_stt

Domaine:

natural language processing

Type de record:

software
Créateur:
Liv
Hôte:
XTrad-STT: A modular, scalable, and open-source Speech-to-Text module supporting low-resource languages. # XTrad-TTS **XTrad-TTS** is a modular, extensible, and open-source Text-to-Speech (TTS) framework designed for the **LiveTrad-S2S** project. It supports multiple backends (e.g., MeloTTS, OpenVoice, Bark, Piper) and is optimized for multilingual applications, including African languages like Swahili. Licensed under **Apache 2.0**, XTrad-TTS is user-friendly with a CLI and Streamlit interface, deployable locally, in the cloud, or on edge devices, and welcomes community contributions. ## Features - **Modular Architecture**: Pluggable TTS backends via a factory pattern. - **Multilingual Support**: Optimized for low-resource languages (e.g., Swahili) with fine-tuning capabilities. - **Real-Time and Batch**: Supports streaming and batch synthesis. - **User-Friendly**: CLI for simple usage and Streamlit for a graphical interface. - **Scalability**: Local (CPU/GPU), cloud (Docker), and edge (low-resource) deployments. - **Open-Source**: Apache 2.0 license, hosted on GitHub and PyPI. ## Installation ```bash pip install xtrad-tts ``` ## Quick Start ```bash xtrad-tts --input text.txt --language sw --output audio.wav ``` ## Streamlit Demo ```bash poetry run streamlit run src/xtrad_tts/ui/app.py ``` ## Contributing See docs/contributing.md for guidelines on adding backends, fine-tuning, or improving documentation. ## License Apache 2.0