The Summarizer 3000: It's better than 2000

A Flask app to summarize PDF articles. This is a final project for a course I am in. I am not a professional. You probably shouldn't use this code.

Purpose

Many students entering University are often overwhelmed by learning how to read academic articles. This Flask app allows users to upload a PDF file of an article or chapter of a book and have GPT-3.5 summarize the main points of the article.

Current issues

• Need to add an auto-return page if a PDF that is too long is submitted. Currently kicks to an internal server error, see above.

The code implies I need to change the limit for tokens...so perhaps the chunking is not working entirely right:

{'message': "This model's maximum context length is 16385 tokens. However, your messages resulted in 18130 tokens. Please reduce the length of the messages.", 'type': 'invalid_request_error', 'param': 'messages', 'code': 'context_length_exceeded'}}

• Could not get the js script to remove the query url, as when it worked the result did not appear on the site... a limitation of my knowledge of javascript and bothered me too much that I had to leave it alone

• Currently the response does not include line breaks. Need to fix: the result in GET does not include line breaks even though it was specified in the prompt.

• Update css and html

• Some PDFs fail to upload, likely due to being improperly extracted (possibly needs to restrict and remove headers?)

  The reason why is a document may be too long, need to fix per first issue.

Table of Contents

• Prerequisites
• Installation
• Structure
• Usage
• Contributing
• License

Note: These instructions are for those running an Ubuntu/Debian-derived system. For those running Windows, Mac, or another operating system the commands may be different.

Prerequisites

sudo apt install python3
python3 -m ensurepip --upgrade 

This installs python (python3) and pip.

You will also require an OpenAI API key. Visit OpenAI for pricing details.

Place the API key in your .zshrc or .bashrc file

echo 'export OPENAI_API_KEY="$KEY"' >> .zshrc
source .zshrc

Installation

Clone and enter the repository

git clone https://github.com/quiltedstars/summarizer_3000 && cd summarizer_3000

Create a .env file with Flask information and optionally your API key

cat <<EOT >> .env
FLASK_APP=app
FLASK_ENV=development

# Once you add your API key below, make sure to not share it with anyone! The API key should remain private.
OPENAI_API_KEY="$KEY"

EOT

Install python requirements

pip install -r requirements.txt

Usage and Structure

To run the app

export FLASK_APP=app.py
flask run

Click the link that follows "*Running on" in order to access the local server and to view the app.

Contributing

1. Fork the repository

2. Create a new branch: git checkout -b $MY_BRANCH

This creates a new branch in your own github to create your own changes

3. Create changes.

4. Push your branch: git push -u origin $MY_BRANCH

This pushes your branch to the original branch

5. Create a pull request

This allows the developer/owner of the original branch to publish the changes you've made

License

MIT License