Audio Transcription

This repository contains utility functions to assist with transcribing audio recordings. The audio recordings are interviews conducted as part of the MSE 343 Human-Computer Interaction (HCI) project. I used a combination of OpenAI's LLMs and Whisper Models as well as HuggingFace's open-source alternatives to allow for variation in models depending on the need.

Table of Contents

• Audio Transcription
  • Table of Contents
  • Prerequisites
  • Virtual Environment Setup
    • Creating a Virtual Environment
    • Activating the Virtual Environment
    • Installing Dependencies
    • Deactivating the Virtual Environment
  • Managing Dependencies with requirements.txt
    • Creating a requirements.txt File
    • Installing from requirements.txt

Prerequisites

• Python 3.x installed on your machine
• Basic knowledge of command-line operations

Virtual Environment Setup

Using a virtual environment is recommended to manage project-specific dependencies and avoid conflicts with other Python projects.

Creating a Virtual Environment

1. Navigate to the project directory:

   cd path/to/Audio_Transcription

2. Create the virtual environment:

   python3 -m venv .venv

   • This command creates a folder named .venv in your project directory.
   • The .venv folder contains a standalone Python installation along with its own pip.

Activating the Virtual Environment

• On Windows:

  .venv\Scripts\activate

• On macOS/Linux:

  source .venv/bin/activate

  • After activation, your command prompt will be prefixed with (.venv) to indicate that the virtual environment is active.

Installing Dependencies

With the virtual environment activated, you can now install project dependencies.

• If you have a requirements.txt file:

  pip install -r requirements.txt

• To install packages individually (e.g., numpy, pandas):

  pip install package_name

Deactivating the Virtual Environment

When you're done working on the project, deactivate the virtual environment:

deactivate

• This returns you to your system's default Python interpreter.

Managing Dependencies with requirements.txt

A requirements.txt file lists all the Python packages needed to run the project, along with their versions.

Creating a requirements.txt File

1. Ensure all necessary packages are installed in your virtual environment.

2. Generate the requirements.txt file:

   pip freeze > requirements.txt

   • This command writes all the installed packages and their versions to requirements.txt.

Installing from requirements.txt

To install all the dependencies listed in requirements.txt:

1. Activate the virtual environment (if not already active).

2. Run the installation command:

   pip install -r requirements.txt

   • This installs all packages specified in the file, ensuring a consistent environment across different setups.

---

Feel free to contribute to this project by submitting issues or pull requests. For any questions or suggestions, please contact the repository maintainer.