This Python application automates the process of generating meeting minutes from an audio recording. It uses the Whisper library for transcription and the OpenAI GPT models for summarizing content, then outputs the result in a Word document.
- Audio File Check: Ensures the presence of an MP3 audio file before processing.
- Audio Transcription: Uses Whisper to accurately transcribe audio.
- Customizable Summary Generation: Leverages OpenAI’s GPT models to generate summaries, with the ability to customize the summary prompt based on user preferences.
- Document Output: Automatically saves the summary in a Word document format.
- Python 3.8 or higher
- Required packages:
whisper
,openai
,python-docx
To get started, clone the repository to your local machine using the command line:
git clone https://github.com/Op27/meeting_minutes_generator.git
This will create a directory named meeting_minutes_generator
which contains all the project files.
After cloning, move into the project directory with:
cd meeting_minutes_generator
Install the necessary Python libraries using pip:
pip install whisper openai python-docx
- Place your MP3 audio file in the root directory of the project.
- Set your OpenAI API key in the script.
- Update the
prompt.txt
file to define the design and style of your meeting minutes. This customization allows the summary generation to align with specific formatting or thematic preferences.
Execute the script by running:
python main.py
The application allows for customization of the summary prompt to cater to specific user needs. This feature enables users to influence the style and focus of the generated summaries, making the tool adaptable for different types of meetings.
This project is open-sourced under the MIT License. You can view the full license text here.
Contributions are welcome! If you'd like to improve the application or suggest features:
- Fork the repository.
- Create a new branch for your changes.
- Make your changes.
- Submit a pull request.