This project implements a text summarizer using Facebook's powerful BART-large-CNN model in a PyTorch framework. The project focuses on providing an efficient and effective solution for summarizing textual content.

• Facebook/bart-large-cnn model: Utilizes the advanced BART-large-CNN model for accurate summarization.
• PyTorch framework: Leverages the flexibility and efficiency of the PyTorch library.
• Modular structure: Organized into distinct pipelines for data ingestion, validation, transformation, training, and evaluation.
• Web application: Provides a user-friendly web interface for text summarization.
• Parameter configuration: Allows customizing the summarization process through adjustable parameters.

1. Clone the repository:

git clone https://github.com/kzebibi/text_summarizer.git
cd text_summarizer

2. Create a virtual environment (recommended):

python -m venv venv
source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

3. Install required packages:

pip install -r requirements.txt
pip install --upgrade accelerate
pip uninstall -y transformers accelerate
pip install transformers accelerate
pip uninstall -y fsspec
pip install fsspec==2023.6.0

Ensure you have datasets==2.10.1 installed.

1. Run the summarization pipeline:

python main.py

This command will execute the following steps:

• Data ingestion
• Data validation
• Data transformation
• Model training
• Model evaluation

2. Start the web application:

python app.py

This will launch a web server, allowing you to input text and generate summaries.

This project is licensed under the MIT License. See the LICENSE file for details.

This project leverages Facebook's BART-large-CNN model and the PyTorch framework. The code structure and implementation draw inspiration from the provided code snippet.