README.md

metadata

title: KittenTTS - High Quality Text-to-Speech
emoji: 🎤
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 5.41.1
app_file: app_minimal.py
pinned: false
license: mit

🎤 KittenTTS - High Quality Text-to-Speech

A Hugging Face Space showcasing the KittenTTS model for high-quality text-to-speech generation.

🚀 Features

• 8 Different Voices: 4 male and 4 female voices to choose from
• High Quality Audio: 24kHz sample rate for crisp, clear speech
• GPU-Free: Works without requiring a GPU
• Easy-to-Use Interface: Simple and intuitive Gradio web interface
• Real-time Generation: Fast speech synthesis with progress tracking

🎵 Available Voices

Voice ID	Gender	Description
expr-voice-2-m	Male	Male voice variant 2
expr-voice-2-f	Female	Female voice variant 2
expr-voice-3-m	Male	Male voice variant 3
expr-voice-3-f	Female	Female voice variant 3
expr-voice-4-m	Male	Male voice variant 4
expr-voice-4-f	Female	Female voice variant 4
expr-voice-5-m	Male	Male voice variant 5
expr-voice-5-f	Female	Female voice variant 5

🛠️ Usage

1. Enter Text: Type or paste your text in the input box
2. Select Voice: Choose from the dropdown menu of available voices
3. Generate: Click the "Generate Speech" button or press Enter
4. Download: Play the generated audio or download it

💻 Technical Details

• Model: KittenML/kitten-tts-nano-0.1
• Sample Rate: 24kHz
• Framework: KittenTTS
• Interface: Gradio
• Audio Format: WAV (24kHz, mono)

🔧 Local Development

To run this locally:

# Clone the repository
git clone <your-repo-url>
cd <your-repo-name>

# Install dependencies
pip install -r requirements.txt

# Run the application
python app.py

📦 Dependencies

• gradio>=4.0.0 - Web interface
• kittentts - TTS framework
• soundfile - Audio file handling
• numpy - Numerical operations
• torch - PyTorch backend
• torchaudio - Audio processing
• transformers - Hugging Face transformers
• accelerate - Model acceleration

🤝 Contributing

Feel free to contribute by:

• Reporting bugs
• Suggesting new features
• Improving the UI
• Adding more voice options

📄 License

This project uses the KittenTTS model. Please refer to the original model's license for usage terms.

🙏 Acknowledgments

• KittenML for the TTS model
• Hugging Face for the Spaces platform
• Gradio for the web interface framework

---

Note: This is a demonstration of the KittenTTS model. For production use, please ensure compliance with the model's license and terms of use.