Real-Time Voice Assistant with FastRTC

This project demonstrates how to create a real-time voice assistant using the FastRTC library in Python. The assistant runs entirely on a local CPU and utilizes local models for speech-to-text and text-to-speech, ensuring privacy and low latency.

Required URLs

• FastRTC Library: https://fastrtc.org/
• Ollama (for running local models): https://ollama.com/
• Python: https://www.python.org/

Setup and Commands

Here are the commands used in the video to set up and run the voice assistant.

1. Install uv (a fast Python package installer and resolver):

   pip install uv

2. Verify the uv installation:

   uv --version

3. Navigate to your project directory:

   cd path/to/your/project

4. Create a virtual environment using uv:

   uv venv

5. Activate the virtual environment:

   • On Windows:

     .venv\Scripts\activate

   • On macOS and Linux:

     source .venv/bin/activate

6. Run the application: This command will automatically install all required dependencies listed in your pyproject.toml file and run the script.

   uv run python code.py

7. Download a model with Ollama: You can download different models to use with the assistant.

   ollama pull <model_name>

   Example used in the video:

   ollama pull gemma:2b

8. List all downloaded Ollama models:

   ollama list