SafePlace

To make the internet a Safer Place.

A simple content moderation project built with Mistral models for assessing text content across several categories:

• sexual content
• hate and discrimination
• violence and threats
• dangerous and criminal content
• selfharm content
• health advice
• financial advice
• law advice
• PII: Personal Identifiable Information

Project Structure

The project is divided into two folders:

• Frontend: Contains the user interface and client-side logic.
• Backend: Contains server-side logic, API endpoints, and integration with Mistral models.

---

Screenshots

Screenshot of the UI.

Screenshot showing moderation results.

---

Setup and Installation

Prerequisites

Install dependencies:

pip install -r requirements.txt

Backend

1. Navigate to the backend folder:

   cd backend

2. Set up environment variables:

   • Create a .env file with the necessary API key.

   MISTRAL_API_KEY=***************
   

3. Run the backend server:

   python -m uvicorn main:app --host 127.0.0.1 --port 8001 --reload

Frontend

1. Navigate to the frontend folder:

   cd frontend

2. Start the frontend server:

   streamlit run main.py

Open the frontend interface in your browser: http://localhost:8501