sophie2chance2 / gpt-note-taker

Data Science 201 Notetaker

GPT Note Taker

Introduction

The GPT Note Taker is a tailored notebook designed for efficient note-taking for UC Berkeley's Data Science 201 course, leveraging the power of OpenAI's GPT models.

I discussed the creation/usage of this notebook with my lecturer, he approved of it and encouraged that I share it with others. However, the use of GPT in the context of education can be controversial, so please use this notebook at your own discretion.

Requirements

• Python 3
• External Python libraries:
  • PyPDF2: For PDF manipulation.
  • openai: To interface with the OpenAI API.
  • IPython: For displaying outputs in a notebook format.

To install these libraries use the following command:

pip install -r requirements.txt

Setup & Instructions

1. Register at OpenAI and create a ChatGPT API key here.
2. Add a credit card.
   Each API call costs a few cents. (Example: Running 7 transcripts and testing costs about 5 cents total.)
3. Create a keys.json file in the root directory and input your OpenAI keys.
   Format of keys.json file:

{
"organization_key": "org-YOUR-KEY-HERE",
"api_key": "sk-YOUR-KEY-HERE"
}

4. Rename your PDF file appropriately.
5. Update the prompt section to match your desired note-taking format.
6. Execute the notebook.

Usage

Example Prompt

f"""Topic: {section.title}
    Detail Level: I would like something between high-level summary and detailed notes.
    Learning Objectives: {'; '.join(section.lo)}
    Format: I prefer the notes to be in bullet-point or table format.
    Course Transcript: {section.text}
    Purpose: The notes are for my personal understanding and to facilitate in class discussion with my peers."""

I asked ChatGPT what inputs it needs to be successful in creating the notes. Then I modified it slightly to fit my needs. I would recommend modifying the prompt to fit your needs as well.

Example Output

Disclaimer

While this tool can enhance your note-taking experience, it's crucial to use it responsibly, especially in academic settings. Ensure you adhere to all academic integrity guidelines and any other applicable ethics principles when utilizing this notebook.

Contributing

Feel free to open issues for bugs or suggestions. If you want to contribute directly, open a pull request with your changes.