PII Data Detection

This repository contains the source code and resources for an NER project aimed at detecting PII Data. The project is based on the Kaggle competition and utilizes a fine-tuned DeBERTa V3 base model to achieve its goal.

Project Structure

• data/: Contains training and test datasets in json format. The training and test set provided are placeholders and should be replaced for actual use.
• model_checkpoints/: Stores trained models' checkpoints.
• EDA.ipynb: Jupyter notebook for exploratory data analysis on the training set.
• model_training.py: Fine-tunes the pre-trained DeBERTa V3 base model on the training set and saves the checkpoint to the model_checkpoints/ folder.
• inference.py: Loads fine-tuned DeBERTa model to make predictions on the test set.

Setup

1. Clone the Repository:

   git clone https://github.com/q-xZzz/pii-data-detection.git
   

2. Install Dependencies:

   pip install -r requirements.txt
   

Running the Project

To get the project up and running, follow these steps:

• (optional)Exploratory Data Analysis: Open EDA.ipynb with Jupyter Notebook or JupyterLab to explore the training dataset.
• Model Training:
  • Run python model_training.py.
• Inference:
  • Run python inference.py to load the model checkpoints and use it to make predictions on the test set.