Skip to content

A-Nama/YAPMA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

32 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Yapma 🎯

Basic Details

Team Name: Delulu

Team Members

  • Member 1: Aisha Nama - Cusat
  • Member 2: Anjana Sankar - Cusat
  • Member 3: Majida Nasrin - Cusat

Hosted Project Link

[mention your project hosted project link here]

Project Description

YAPMA tackles the critical challenge of preventing harmful AI misuse by stopping unethical prompts at the source. We're shaping a safer, more responsible AI future—because it matters.

The Problem statement

We're solving the absurd reality where people misuse AI to create harmful and unethical content—because honestly, who thought making AI a supervillian was a good idea? 😤 YAPMA steps in to say, "Not today, bestie!"

The Solution

We’re giving AI a moral compass and a Gen Z vibe! 🌟 YAPMA sniffs out shady prompts, serves up sassy warnings, and stops harmful content before it even exists. Think of us as that loyal bestie your AI didn’t know it needed!

Technical Details

Technologies/Components Used

For Software:

  • Python
  • Streamlit, FastAPI
  • Google-generativeai, Python-dotenv, Requests, Transformers
  • Render

Implementation

For Software:

Installation

pip install -r requirements.txt

Run

uvicorn app.main:app --reload

streamlit run frontend\app.py

Project Documentation

For Software:

Screenshots

Screenshot1 Home page of YAPMA

Screenshot2 YAPMA dealing with unacceptable comments

Screenshot3 YAPMA dealing with acceptable comments

Diagrams

Workflow User inputs flow from Streamlit UI to FastAPI backend, where DistilBERT classifies prompts as Acceptable or Unacceptable, with GenZ feedback according to it.

Project Demo

Video

[https://drive.google.com/file/d/1cFuntqj6_IhYutVYxijsEt1OIjSstrj1/view?usp=sharing] The video demonstrates how YAPMA classifies the prompts as Acceptable and Unacceptablewith a certain confidence level.

Team Contributions

  • Aisha Nama: Backend development, integrated pre-trained DistilBERT model for prompt moderation, version control.
  • Anjana Sankar: Backend development, integrated Gemini API for GenZ feedback generation, handled hosting.
  • Majida Nasrin: UI/UX design, Streamlit frontend development, created content and media for documentation.

Made with ❤️ by Delulu

Releases

No releases published

Packages

No packages published

Languages