- Home
- Text-to-Speech Software
- TensorFlowTTS
On the recordRead 13 Oct 202620 facts
TensorFlowTTS
On the record
- Architectures
- Supported architectures include Tacotron-2, MelGAN, Multiband-MelGAN, FastSpeech, FastSpeech 2, Parallel WaveGAN, and HiFi-GAN.github.com · read 13 Oct 2026
- Commercial use
- Yesgithub.com · read 20 Sep 2026
- Data compatibility
- Its preprocessing is based on ESPnet, and the README says the resulting data can be combined with models from the ESPnet repository.github.com · read 13 Oct 2026
- Datasets
- The documented preprocessing currently supports LJSpeech, KSS, Baker, LibriTTS, Thorsten, and Synpaflex datasets.github.com · read 12 Oct 2026
- Deployment
- The project supports TFLite conversion, C++ inference, and deployment to mobile devices or embedded systems.github.com · read 13 Oct 2026
- Export formats
- WAVgithub.com · read 20 Sep 2026
- Installation
- The project documents installation with pip or from its GitHub source repository.github.com · read 12 Oct 2026
- Integrations
- The project documents integration with Hugging Face Spaces and Gradio, and hosting pretrained models on the Hugging Face Hub.github.com · read 13 Oct 2026
- Languages
- The README lists Chinese, Korean, English, French, and German support, and says the framework can be adapted for other languages.github.com · read 13 Oct 2026
- License
- The repository states that its models are licensed under Apache 2.0.github.com · read 13 Oct 2026
- Mobile examples
- The README includes an Android example and reports iOS support for FastSpeech 2 and MB MelGAN.github.com · read 13 Oct 2026
- Model reuse
- Pretrained models can be loaded from Hugging Face Hub using the project’s inference API.github.com · read 12 Oct 2026
- Models
- It provides Tacotron-2, MelGAN, Multiband-MelGAN, FastSpeech, FastSpeech 2, Parallel WaveGAN, and HiFi-GAN architectures.github.com · read 12 Oct 2026
- Purpose
- TensorFlowTTS provides real-time speech synthesis architectures built on TensorFlow 2.github.com · read 13 Oct 2026
- Requirements
- The repository says it was tested on Ubuntu 18.04 with Python 3.7 or later and TensorFlow versions 2.2 through 2.6.github.com · read 13 Oct 2026
- Support
- The README provides project contact email addresses and links to tutorials, notebooks, and Colab examples.github.com · read 13 Oct 2026
- Supported datasets
- The README lists ljspeech, kss, baker, libritts, thorsten, and synpaflex for preprocessing, with reformatting instructions for some datasets.github.com · read 13 Oct 2026
- Training
- Features include mixed-precision training and single- or multi-GPU training and gradient accumulation.github.com · read 13 Oct 2026
