HUGGINGFACE_DEPLOYMENT.md

🚀 Hugging Face Spaces Deployment Guide

📁 Files Created for Docker Deployment:

• Dockerfile - Main Docker configuration
• requirements-docker.txt - Optimized dependencies for Docker
• .dockerignore - Excludes unnecessary files from build
• start-hf.sh - Startup script for Hugging Face Spaces
• README.md - Hugging Face Spaces metadata

🔧 Deployment Steps:

1. Create New Hugging Face Space

1. Go to Hugging Face Spaces
2. Click "Create new Space"
3. Choose:
   • Space name: marine-guard-api
   • License: MIT
   • SDK: Docker
   • Hardware: CPU basic (free tier)

2. Upload Backend Files

Upload these files to your Hugging Face Space:

Dockerfile
requirements-docker.txt
start-hf.sh
README.md
app/
├── __init__.py
├── main.py
├── config.py
├── database.py
├── dependencies.py
├── schemas.py
├── core_security.py
├── routers/
│   ├── __init__.py
│   ├── auth.py
│   └── incidents.py
└── services/
    ├── __init__.py
    ├── incidents.py
    ├── ml_model.py
    ├── nlp.py
    └── users.py
models/
├── threat_model.pkl
└── severity_model.pkl

3. Set Environment Variables in Hugging Face

In your Space settings, add these environment variables:

MONGODB_URI=mongodb+srv://deepdblm_db_user:IqLKnKhwLLSOP1Ka@cluster0.0u1vpow.mongodb.net/?retryWrites=true&w=majority&appName=Cluster0
JWT_SECRET_KEY=your-secure-secret-key-here
ALLOWED_ORIGINS=https://marine-pollution-detection.vercel.app,http://localhost:3000

4. Update Frontend Configuration

Once deployed, update your frontend .env file:

VITE_API_BASE_URL=https://your-username-marine-guard-api.hf.space/api

🎯 Expected URLs:

• API Base: https://your-username-marine-guard-api.hf.space
• Health Check: https://your-username-marine-guard-api.hf.space/health
• Auth Endpoint: https://your-username-marine-guard-api.hf.space/api/auth/login

🔍 Troubleshooting:

• Check the Logs tab in your Hugging Face Space for any errors
• Ensure all environment variables are set correctly
• The space will take 2-3 minutes to build and start
• Models will be loaded automatically on startup

📊 Features Included:

• ✅ FastAPI backend with all endpoints
• ✅ MongoDB Atlas connection
• ✅ JWT authentication
• ✅ ML model inference (threat & severity classification)
• ✅ CORS configured for Vercel frontend
• ✅ Health check endpoint
• ✅ Automatic model loading on startup