Skip to content

Latest commit

 

History

History
112 lines (72 loc) · 4.23 KB

README.md

File metadata and controls

112 lines (72 loc) · 4.23 KB

Orcus_Banner

Observational Recognition of Content with Unnatural Speech.

Kuberwastaken - TREAT Version 2.0 License Apache 2.0

Named after the Roman god of the underworld and the punisher of broken oaths, ORCUS is dedicated to upholding integrity and authenticity in the digital realm by flagging AI generated content disguised as normal comments.

Watch ORCUS in action on LinkedIn

Mission

ORCUS' mission is to address the growing issue of AI-generated content, particularly on platforms like LinkedIn, where AI generated comments fill up the discussion without adding significantly to the original post or conversation.

These automated responses often overshadow genuine human engagement, contributing to a superficial online environment that diminishes the quality of discourse.

ORCUS strives to foster a cultural shift. By encouraging digital literacy and active participation, the project aims to reshape online spaces, restore trust, and ensure that social media remains a platform for thoughtful, original interactions where human creativity and insight are celebrated over automated, generic outputs, without much thought to it.

Features

  • AI Detection Algorithms: ORCUS uses the highly advanced and Open Source Roberta-Base-OpenAI-Detector Model from Hugging Face

  • Interactive Interface: A user-friendly interface based on React allows users to input text and analyze its authenticity.

  • Real-Time Feedback: Provides almost immediate insights and suggestions for creating more authentic and engaging content.

  • Custom GPT-2 Generated Messages: ORCUS uses GPT-2 to dynamically generate engaging and creative AI detection messages with emojis, adding a unique flair to every alert and detection.

  • Community-Driven: Encourages a community of users committed to promoting originality and discouraging the overuse of AI for trivial purposes.

Installation Instructions

Prerequisites

  • Star the Repository to Show Your Support :P

  • Clone the Repository to Your Local Machine:

    git clone https://github.com/Kuberwastaken/ORCUS.git

Hugging Face Instructions

  1. Login to Hugging Face in Your Environment:

    Run the following command in your terminal:

    huggingface-cli login

    Enter your Hugging Face access token when prompted.

  2. Download the Roberta-Base-Openai-Detector and GPT-2 Models:

    The models will be downloaded automatically when running the script analysis for the first time.

Environment Setup

To set up the development environment, you will need to create a virtual environment and install the necessary dependencies.

  1. Create a Virtual Environment:

    python3 -m venv orcus

  2. Activate the Virtual Environment:

    source treat-env/bin/activate   # On Unix or MacOS
treat-env\Scripts\activate      # On Windows

  3. Install Dependencies:

    Navigate to the project directory and run:

    pip install -r requirements.txt

Project Usage

  1. Start the interface and the model

    npm start

  2. Analyze Scripts:

    You would be taken to a localhost website where you can manually enter a comment or text in the provided text area and click "Analyze Script."

Technologies Used

  • Frontend: React

  • Backend: Node.js

  • Machine Learning: TensorFlow.js

License

This project is licensed under the Apache 2.0 License. See the LICENSE file for details.

Contribution

Contributions are welcome and highly encouraged! Please fork the repository and submit a pull request for review!