Skip to content

Latest commit

 

History

History
113 lines (86 loc) · 4.42 KB

README.md

File metadata and controls

113 lines (86 loc) · 4.42 KB

📢 Indic TTS Models - Open Source Repository

🎙️ Overview

This repository provides an open-source implementation of various Indic Text-to-Speech (TTS) models that support multiple Indian languages. The models enable text input conversion into natural-sounding speech across various Indian languages. This project aims to enhance accessibility and voice synthesis research.

🚀 Why This Repository?

  • Supports multiple Indian languages
  • Offers easy-to-use Colab notebooks
  • Provides open-source solutions for research and development
  • Encourages community contributions to improve Indic TTS

🔥 Available Models

1️⃣ ai4bharat/indic-parler-tts

  • Size: 3.8 GB
  • Supported Languages: 20 Indic languages
  • Inputs: Text, Voice description
  • Output: Audio
  • License: Apache 2.0 (Free for all uses)
  • Colab Notebook: Try Here

📝 Results: ✔️ Very natural and clear audio
✔️ Good accent
❌ Large model size
❌ Some inconsistencies

🎧 Sample Output:

2️⃣ facebook/mms-tts-hin / facebook/mms-tts-guj

  • Size: 150-160 MB
  • Supported Languages: Hindi (Separate models for other languages)
  • Inputs: Text
  • Output: Audio
  • License: CC-BY-NC 4.0 (Non-commercial use only)
  • Colab Notebook: Try Here
  • Other Languages: Find More Models

📝 Results: ✔️ Natural and clear audio
✔️ Small model size
❌ Requires switching models for different languages
❌ Some voice inconsistencies

🎧 Sample Output:
🔗 Listen Here

3️⃣ ai4bharat/vits_rasa_13

  • Size: 170 MB
  • Supported Languages: Assamese, Bengali, Bodo, Dogri, Kannada, Maithili, Malayalam, Marathi, Nepali, Punjabi, Sanskrit, Tamil, Telugu
  • Inputs: Text
  • Output: Audio
  • License: Apache 2.0 (Free for all uses)
  • Colab Notebook: Try Here

📝 Results: ✔️ Clear audio
✔️ Mixed accent support
❌ Slight tone inconsistencies

🎧 Sample Output:
🔗 Listen Sample 1
🔗 Listen Sample 2

🚀 How to Use

1️⃣ Open the Colab notebook for the model you want to use.
2️⃣ Run the provided cells to install dependencies and set up the environment.
3️⃣ Input text (and voice description, if applicable).
4️⃣ Generate speech and listen to the output!

🛠️ Contributing

We welcome contributions from the community! If you’d like to contribute:
📌 Bug Reports & Issues: Open an issue if you find any bugs.
📌 Feature Requests: Suggest new features via discussions or pull requests.
📌 Pull Requests: Fork the repo, make changes, and submit a pull request!

Contribution Guidelines:

  • Ensure your changes follow best practices.
  • Keep the repository structured and well-documented.
  • Respect the licenses of individual models.

📜 License

This repository includes models with different licenses:

  • ai4bharat models: Apache 2.0 (Permissive, allows commercial use)
  • Facebook MMS-TTS models: CC-BY-NC 4.0 (For non-commercial use only)

📌 Note: If you plan to use any model commercially, verify its license compliance.

💡 Acknowledgments

Special thanks to AI4Bharat and Facebook Research for open-sourcing these Indic TTS models and contributing to speech synthesis research.

🌍 Join Our Community!

If you're working on Indic TTS, we'd love to collaborate!
💬 Join the Discussion: GitHub Discussions
🐦 Follow us on Twitter: @YourTwitterHandle
📢 Stay Updated: Watch the repo for new models and updates!