README.md

---

Voice-Activated OpenAI GPT-4 Assistant

This project is a voice-activated assistant using OpenAI's GPT-4. It records the user's voice, transcribes it, feeds it to the GPT-4 model, and plays back the response using text-to-speech. It also measures the time taken for each step in the process.

Setup and Usage

Prerequisites

• Python 3.8+
• OpenAI API key
• Required Python packages: PyQt5, sounddevice, numpy, openai, gTTS, pydub, scipy

Installation

1. Clone the repository:

   git clone https://github.com/mh-nyris/openAI.git
   

2. Navigate into the project directory:

   cd openAI
   

3. Install the necessary packages:

   pip install -r requirements.txt
   

   Note: It's recommended to do this in a virtual environment to avoid installing packages globally.

Additional Installations

This project also depends on a few packages that need to be installed via Homebrew:

brew install ffmpeg

ffmpeg is required for pydub to work correctly.

Usage

1. Replace "placeholder" in the following lines of code with your own OpenAI API credentials:

   openai.organization = "placeholder"
   openai.api_key = 'placeholder'  # Replace with your OpenAI key

2. Run the script:

   python voice-communication-gpt4.py
   

3. Interact with the application GUI to start and end voice recording.

License

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