convoClone : AI clone trained on Whatsapp Chats

This project demonstrates an AI assistant with WhatsApp integration for context-aware responses using Twilio and a Llama model fine tuned on Whatsapp chats.

Features

• Real-time WhatsApp message processing
• Context-aware responses
• Flask API for predictions

Setup

1. Clone the repository.
2. Install dependencies:

   pip install -r requirements.txt

3. Set environment variables:

   export TWILIO_ACCOUNT_SID="your_account_sid"
   export TWILIO_AUTH_TOKEN="your_auth_token"
   export TWILIO_NUMBER="your_twilio_number"

4. Run the application:

   python api.py

Usage

• Send messages via WhatsApp to interact with the assistant.
• Use the /predict endpoint for API-based predictions.

File Structure

• api.py: Defines API routes.
• model.py: Handles model inference.
• utils.py: Includes utility functions.
• config.py: Configuration settings.
• requirements.txt: Python dependencies.
• README.md: Project overview and instructions.