README.md

metadata

title: Mental Wellness Assistant
emoji: 🧠
colorFrom: indigo
colorTo: purple
sdk: gradio
sdk_version: 5.8.0
app_file: app.py
pinned: false
license: mit

Mental Wellness Assistant

A sophisticated AI-powered mental wellness assistant using state-of-the-art Mistral-7B language models. This application provides both therapeutic conversation and mindfulness guidance through an intuitive chat interface.

Features

1. Dual-Agent System

• Therapeutic Agent: Provides empathetic conversations and emotional support
• Mindfulness Agent: Offers guided meditation and mindfulness exercises

2. Advanced AI Capabilities

• Powered by Mistral-7B-Instruct-v0.2
• Context-aware responses
• Natural conversation flow

3. User Interface

• Clean, intuitive chat interface
• Easy switching between therapeutic and mindfulness modes
• Example prompts for quick start

Technical Details

Requirements

• Python 3.8+
• CUDA-compatible GPU recommended
• 16GB+ RAM recommended

Key Dependencies

• Gradio: Web interface
• Transformers: Hugging Face model support
• PyTorch: Deep learning framework
• Hugging Face Hub: Model hosting and management

Model Information

• Base Model: Mistral-7B-Instruct-v0.2
• Optimization: 16-bit precision for efficiency
• Auto device mapping for optimal resource usage

Usage

1. Mode Selection

   • Choose between "Therapeutic Mode" and "Mindfulness Mode"
   • Each mode provides specialized responses for different needs

2. Conversation

   • Type your message in the chat interface
   • Get AI-generated responses based on the selected mode
   • Use example prompts for guidance

3. Example Interactions

   • "I've been feeling anxious lately"
   • "Guide me through a breathing exercise"
   • "I need help managing stress"
   • "Can you teach me meditation?"

Important Disclaimers

1. Not a Medical Service

   • This is a demonstration project
   • Not a substitute for professional mental health care
   • Seek professional help for serious concerns

2. Privacy Notice

   • Conversations are not stored permanently
   • Data is processed securely
   • Use discretion when sharing personal information

Emergency Resources

If you're experiencing a mental health crisis:

• Emergency Services: 911 (US)
• National Crisis Hotline: 988
• Crisis Text Line: Text HOME to 741741

Development

Local Setup

1. Clone the repository
2. Install dependencies: pip install -r requirements.txt
3. Set up HF_TOKEN in environment variables
4. Run: python app.py

Hugging Face Spaces

This application is deployed on Hugging Face Spaces and uses:

• Gradio SDK
• Environment variables for secure token management
• Automated logging for monitoring

License

MIT License - Feel free to use and modify with attribution

Acknowledgments

• Mistral AI for the base model
• Hugging Face for model hosting and tools
• Gradio team for the interface framework