ABKUMA01/BE500Project1

Article Summarizer via facebook/bart-large-cnn

This is an interactive Gradio web app that can summarize articles and paragraphs concisely and accurately. It uses a pretrained model, facebook/bart-large-cnn, from the Hugging Face Hub to do the text summarization.

Objective

The goal of this web app is to demonstrate how to use a pretrained transformer model into a user-friendly Gradio interface and deploy it on Hugging Face Spaces.

Model Information

• —Model Name: facebook/bart-large-cnn
• —Task: Text summarization
• —Framework: Hugging Face Transformers
• —Training Data: abisee/cnn_dailymail
• —License: mit

Run Locally

1. 1.Copy this repository or download the files.
2. 2.Install the required dependencies: pip install -r requirements.txt
3. 3.Run the app: python app.py
4. 4.Click on the url displayed in the terminal

Limitations

• —The summarizer may incorrectly/inaccurately rephrase sentences for longer documents.
• —BART's context window is limited to ~1024 tokens.

Credits

• —Model: facebook/bart-large-cnn (Hugging Face)
• —Interface: Gradio
• —Author: Arjun Kumar