📰 News Summarization using T5 Model

This project implements a news summarization system using the T5-base model, fine-tuned on the CNN/DailyMail dataset from Hugging Face. The project covers data pre-processing, exploratory data analysis (EDA), model fine-tuning, and evaluation using ROUGE scores.

---

📚 Dataset Information

Dataset used: CNN/DailyMail

• Train Size: 287,113 samples
• Validation Size: 13,368 samples
• Test Size: 11,490 samples

---

🖥️ Web Dashboard

---

📊 Evaluation Results

Metric	Score
ROUGE-1	0.2969
ROUGE-2	0.1204
ROUGE-L	0.2483
ROUGE-Lsum	0.2483

🛠️ Project Setup

1. Clone the repository:

git clone https://github.com/omkar-79/news-summarization.git
cd news-summarization

2. Create a virtual environment (Optional):

# For Python 3.x
python3 -m venv venv
source venv/bin/activate    # On Linux/Mac
# OR
venv\Scripts\activate       # On Windows

3. Install required packages:

pip install -r requirements.txt

4. Run the Jupyter Notebook:

• Open the .ipynb file using Jupyter Notebook or Jupyter Lab.
• Run all the cells to fine-tune the model.
• The fine-tuned model will be saved to:

./t5_finetuned

5. Run the Flask Application:

python app.py

6. Open the Interface:

After running app.py, open index.html in your browser to access the application.