Skip to content

Latest commit

 

History

History
44 lines (28 loc) · 1.14 KB

README.md

File metadata and controls

44 lines (28 loc) · 1.14 KB
Raw

Machine Generated Text (MGT) Detector UI

This project is a machine-generated text detector that uses multiple existing MGT detection models through the Hugging Face API and text statistics to determine if text is machine-generated or not. It combines the power of various models and statistical analysis to make more accurate predictions regarding the origin of the text.

Getting Started

To analyse the dataset and generate statistics required for the graphs , please follow these steps:

  1. Go to the analysis folder
cd src/analysis
  1. Run analysis.py to generate combined_analysis.json
python3 analysis.py
  1. Run graph.py to verify that the results look valid (Optional)
python3 graph.py
  1. Move combined_analysis.json to the public folder if updated.

To run the user interface, please follow these steps:

  1. Install the dependencies.
pnpm install

  1. Copy .env.local.example to a new file .env.local and add your Hugging Face API key.

  2. Run the development server.

pnpm run dev
  1. Open http://localhost:3000 with your browser to see the UI.