# Question Answering with T5 Transformer

This project implements a question-answering task using the T5 (Text-To-Text Transfer Transformer) model, fine-tuned on the Stanford Question Answering Dataset (SQuAD). The T5 model is capable of understanding and generating human-like text, making it suitable for various NLP tasks, including question answering.

## Features

- Fine-tuning the T5 model on the SQuAD dataset
- Training and evaluation scripts
- Support for loading pre-trained models from Hugging Face
- Metrics and visualizations for model performance

## Dataset

The project uses the SQuAD dataset, a popular benchmark for machine reading comprehension. The dataset can be downloaded automatically using the Hugging Face Datasets library.

## Requirements

- Python 3.7+
- PyTorch
- Hugging Face Transformers
- Hugging Face Datasets
- tqdm
- pandas

To install all dependencies, run:
```bash
pip install -r requirements.txt

Usage

1. Training the Model

   Run the following command to fine-tune the T5 model on the SQuAD dataset:

   python t5_qa_training.py --dataset squad --model t5-small

2. Evaluating the Model

   After training, evaluate the model's performance:

   python evaluate_model.py --model_path output/t5_qa_model

3. Inference

   Use the trained model to answer custom questions:

   python inference.py --question "What is T5?" --context "T5 is a text-to-text transfer transformer model..."

Results

Metric	Score
Exact Match	59.2%
F1 Score	88.4%

Project Structure

• t5_qa_training.py: Script for training the T5 model.
• evaluate_model.py: Script for evaluating the trained model.
• inference.py: Script for performing inference with the trained model.
• README.md: Project description and usage instructions.

License

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

Acknowledgements

• Hugging Face Transformers for providing the pre-trained T5 model.
• SQuAD Dataset for the training data.