BERT-based Text Classification for IMDB Reviews

This project implements a BERT-based text classification model to perform sentiment analysis on IMDB movie reviews. It uses PyTorch and the Hugging Face Transformers library to fine-tune a BERT model for binary classification (positive/negative sentiment).

Project Structure

text-classification/
├── data/
│   └── imdb_data.csv
├── src/
│   ├── data/
│   │   └── dataset.py
│   ├── model/
│   │   └── classifier.py
│   ├── train.py
├── main.py
├── config.yaml
├── requirements.txt
└── README.md

Setup

1. Clone the repository:

   git clone https://github.com/your-username/text-classification.git
   cd text-classification

2. Create and activate a Conda environment:

   conda create --name bert-classifier
   
   conda activate bert-classifier

3. Install the required packages:

   pip install -r requirements.txt

4. Prepare your data:

   • Place your IMDB dataset (CSV format) in the data/ directory.
   • Ensure your CSV file has 'review' and 'sentiment' columns.

5. Configure the project:

   • Edit config.yaml to set your desired parameters, data paths, and model settings.

Usage

To train the model:

python main.py

Features

• BERT-based text classification
• Custom attention pooling mechanism
• Training with early stopping
• Weights & Biases integration for experiment tracking (optional)

Configuration

The config.yaml file contains all the configurable parameters for the project. Key configurations include:

• Data paths
• Model hyperparameters (e.g., learning rate, batch size)
• Training settings (e.g., number of epochs, early stopping threshold)

Contributing

Contributions to this project are welcome! Please fork the repository and submit a pull request with your proposed changes.

Acknowledgments

• The BERT model implementation is based on the Hugging Face Transformers library.
• The IMDB dataset used for this project is publicly available.