Realtime TTS Assistant

This project is a real-time Text-to-Speech (TTS) assistant powered by OpenAI's GPT-3 model and a local inference server. It converts text input into speech output as you type, making it an interactive and hands-free tool.

Prerequisites

Before you begin, ensure you have the following:

• Python 3.x
• Required Python packages (install via pip):
  • openai
  • pyttsx3
  • nltk
• LM Studio

Setup

1. Clone this repository:

   git clone https://github.com/your-username/Realtime-TTS-Assistant.git](https://github.com/ProjCRys/Realtime-TTS-Assisstant.git
   cd Realtime-TTS-Assistant

2. Configure OpenAI API settings or you can use local host for local language models in the script:

   • Set the API base and key in the script:

   openai.api_base = 'http://localhost:1234/v1'  # Update with your API endpoint
   openai.api_key = 'YOUR_API_KEY'  # Replace with your API key or simply leave it empty for local language models

3. Create a context.txt file and provide system prompt for the assistant.

4. Customize prompt format by modifying the prefix and suffix variables as needed.

5. Run the script:

   python lmStudio_to_TTS.py

Usage

1. Run the script and interact with the assistant in your terminal. Type your input, and the assistant will respond with speech output in real-time.

2. To exit the program, type 'exit' and press Enter.

Features

• Real-time text-to-speech conversion.
• Customizable prompt format.
• Sentence-based TTS for natural conversation flow.

Sample Video

https://m.youtube.com/watch?v=TrOsG4jnSNw&t=5s

Contributing

Contributions are welcome! Feel free to open issues or submit pull requests.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• OpenAI for the GPT-3 model.
• Python community and open-source contributors.
• LM Studio for Local Server Inference