DocSummarizer

I built this just to get hands-on experience with Hugging Face models, Streamlit, and Transformers. It's a simple Streamlit app to upload PDF files and generate concise summaries using Hugging Face Transformers.

Features

• Upload PDF files via a simple web interface
• Automatic extraction of text from PDFs
• Summarization using the BART model (facebook/bart-large-cnn)
• Handles long documents by chunking text
• Progress bar for summarization process

Requirements

• Python 3.7+
• Streamlit
• transformers
• torch

Installation

1. Clone the repository:

   git clone <repository-url>
   cd DocSummarizer

2. Install dependencies:

   pip install -r requirements.txt

Usage

1. Run the Streamlit app:

   streamlit run main.py

2. Open the provided local URL in your browser.
3. Upload a PDF file and view the generated summary.

Notes

• The summarization model is loaded locally. Ensure you have downloaded the facebook/bart-large-cnn model or have internet access for the first run.
• GPU support is recommended for faster summarization.