PPGT-GEN is a convenient local web application for using the PagPassGPT model, replacing the complex Command Line Interface (CLI) usage. The application provides a user-friendly interface for:

• Password Generation: Using PagPassGPT model to generate intelligent passwords
• Password Strength Checking: Analyzing and evaluating password security
• Entropy Analysis: Measuring randomness and unpredictability of passwords
• Pattern Checking: Detecting patterns and rules in passwords
• Model Evaluation: Comparing effectiveness of different models

• Python 3.8+
• Flask
• PyTorch
• Transformers
• Other dependencies (see requirements.txt)

git clone https://github.com/WanThinnn/PPGT-GEN.git
cd PPGT-GEN

pip install -r requirements.txt

To obtain the model and dataset for training, please refer to the original project:

PagPassGPT Original Repository

1. Download the preprocessed RockYou dataset from the original project
2. Copy the entire dataset to the directory:

   PPGT-GEN/dataset/
   

   Dataset directory structure:

   dataset/
   ├── rockyou-cleaned.txt
   └── [other dataset files]
   

1. After training the model from the original project, you will have a last-step directory
2. Copy the entire last-step directory to:

   PPGT-GEN/model/last-step/
   

   Model directory structure:

   model/
   └── last-step/
       ├── config.json
       ├── generation_config.json
       ├── optimizer.pt
       ├── pytorch_model.bin
       ├── rng_state.pth
       ├── scheduler.pt
       ├── trainer_state.json
       └── training_args.bin
   

python app.py

The application will run at: http://localhost:5000

• Path: /ppgt_gen
• Function: Generate passwords using PagPassGPT model
• Options: Configure length, patterns, and password generation parameters

• Path: /strength_checker
• Function: Check and evaluate password strength
• Features: Detailed analysis of security factors

• Path: /entropy_checker
• Function: Analyze password entropy (randomness)
• Features: Measure unpredictability

• Path: /pattern_checker
• Function: Detect and analyze patterns in passwords
• Features: Identify common rules and patterns

• Path: /evaluate_checker
• Function: Evaluate and compare model effectiveness
• Features: Statistics and analysis charts

PPGT-GEN/
├── app.py                      # Main Flask application
├── requirements.txt            # Dependencies
├── README.md                   # This documentation
├── patterns.txt               # Pattern files
├── app.log                    # Log file
├── dataset/                   # Dataset directory
│   └── rockyou-cleaned.txt
├── model/                     # Model directory
│   └── last-step/
├── libanalyst/               # Analysis library
│   ├── password_strength_checker.py
│   ├── password_entropy_checker.py
│   ├── password_pattern_checker.py
│   └── password_evaluate_checker.py
├── libppgt/                  # PPGT core library
│   ├── __init__.py
│   ├── concat_pattern_password.py
│   ├── DC-GEN.py
│   ├── get_pattern_rate.py
│   ├── normal-gen.py
│   └── tokenizer/
├── templates/                # HTML templates
│   ├── index.html
│   ├── ppgt_gen.html
│   ├── strength_checker.html
│   ├── entropy_checker.html
│   ├── pattern_checker.html
│   └── evaluate_checker.html
└── static/                   # CSS & JS files
    ├── css/
    └── js/

• GET / - Main page
• GET /ppgt_gen - Password Generator
• GET /strength_checker - Strength Checker
• GET /entropy_checker - Entropy Checker
• GET /pattern_checker - Pattern Checker
• GET /evaluate_checker - Evaluate Checker

• POST /api/generate_password - Generate passwords
• POST /api/check_strength - Check strength
• POST /api/analyze_entropy - Analyze entropy
• POST /api/check_pattern - Check patterns
• POST /api/evaluate_model - Evaluate model

• Lại Quan Thiên - WanThinnn
• Mai Nguyễn Nam Phương - Mai Nguyen Nam Phuong - Cyber Security
• Trần Thế Hữu Phúc - tranthehuuphuc
• Hồ Diệp Huy - hohuyy

• Original Project: PagPassGPT
• Issues: Create an issue on this GitHub repository
• Documentation: See additional documentation in the docs/ folder

• Thanks to the PagPassGPT development team for the original model and dataset
• Thanks to the open source community for contributions and support

Security Note: Only use this application in a local environment and do not share the model or dataset externally without proper permissions.