QUICK_START.md

Quick Start Guide - Intel Image Classifier

Get up and running in minutes!

Prereq: Install Models

Before deploying, you need your trained models. Place them in:

backend/api/models/
├── pytorch_model.pth      # PyTorch model weights
└── model_best.keras       # TensorFlow/Keras model

Note: If you don't have these files yet, see /ml/ directory for training scripts.

---

Option A: Deploy to Hugging Face (Recommended ⭐)

Step 1: Create Space on Hugging Face

1. Go to huggingface.co/spaces
2. Click "Create new Space"
3. Choose:
   • Name: Intel_classification
   • License: MIT
   • Space SDK: Docker
   • Visibility: Public

Step 2: Clone and Update Repository

# Clone this repo
git clone https://github.com/danielle2035/Intel_classification.git
cd Intel_classification

# Add your trained models to backend/api/models/
cp /path/to/pytorch_model.pth backend/api/models/
cp /path/to/model_best.keras backend/api/models/

# Add Hugging Face remote
git remote add hf https://huggingface.co/spaces/YOUR_HF_USERNAME/Intel_classification

Step 3: Deploy!

git push hf main

Done! Watch your Space build and deploy automatically. Access it at:

https://huggingface.co/spaces/YOUR_HF_USERNAME/Intel_classification

---

Option B: Run Locally with Docker

Easiest Way

# Build the image
docker build -t intel-classifier .

# Run it
docker run -p 7860:7860 intel-classifier

Then open: http://localhost:7860

With Docker Compose (Development)

docker-compose up --build

Services:

• Frontend: http://localhost:3000
• Backend: http://localhost:8000
• Docs: http://localhost:8000/swagger

---

Option C: Run Locally Without Docker

Backend Setup

cd backend/api

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

# Install dependencies
pip install -r ../requirements.txt

# Run migrations
python manage.py migrate

# Start server
python manage.py runserver 8000

Keep terminal open. Backend runs on http://localhost:8000

Frontend Setup (New Terminal)

cd frontend

# Install dependencies
npm install

# Start development server
npm start

Frontend runs on http://localhost:3000

---

Testing Your Deployment

1. Check Health

curl http://localhost:7860/health/
# Expected: {"status": "healthy", "service": "Intel Image Classifier API", "version": "1.0.0"}

2. Classify an Image

curl -X POST \
  -F "image=@test_image.jpg" \
  -F "model=pytorch" \
  http://localhost:7860/api/classify/

3. Visit Web Interface

Open in browser: http://localhost:7860

4. Check API Docs

• Swagger: http://localhost:7860/swagger/
• ReDoc: http://localhost:7860/redoc/
• Admin Panel: http://localhost:7860/admin/ (user: admin, pass: admin)

---

Troubleshooting

"Port 7860 already in use"

# Find what's using it
lsof -i :7860

# Kill the process
kill -9 <PID>

"Models not found"

Ensure these files exist:

• backend/api/models/pytorch_model.pth
• backend/api/models/model_best.keras

If missing, only one model will be available.

"CORS Error"

This usually means backend and frontend are on different domains. Verify:

• Docker mode: Both on same domain ✅
• Local dev: Frontend 3000, Backend 8000 - they communicate via proxy ✅
• HF Spaces: Auto-configured ✅

"Models take too long to load"

First startup loads models into memory. This can take 1-2 minutes for large models. Subsequent requests are fast!

---

Common Tasks

Change Confidence Threshold

Edit backend/api/notifications/api_views.py:

CONFIDENCE_THRESHOLD = 0.6  # Change this value

Add Custom Classes

Update CLASSES list in backend/api/notifications/api_views.py:

CLASSES = ["buildings", "forest", "glacier", "mountain", "sea", "street", "YOUR_CLASS"]

Then retrain your models.

Use a Different Model

Add to backend/api/models/:

• pytorch_model.pth
• model_best.keras

The API automatically detects available models.

---

File Structure Reference

intel-classifier/
├── Dockerfile                    # Docker configuration
├── README.md                     # Main documentation
├── DEPLOYMENT.md                # Detailed deployment guide
├── REORGANIZATION.md            # What changed
├── QUICK_START.md              # This file!
│
├── backend/
│   ├── api/notifications/       # Image classification API
│   │   └── api_views.py
│   ├── models/                  # Your trained models
│   │   ├── pytorch_model.pth
│   │   └── model_best.keras
│   └── requirements.txt
│
├── frontend/                    # React web interface
│   ├── src/App.js
│   └── package.json
│
└── ml/                         # Training scripts (for reference)
    └── models/

---

Next Steps

1. ✅ Add your trained models
2. ✅ Test locally (Docker or native)
3. ✅ Push to Hugging Face Spaces
4. ✅ Share with friends!
5. 📊 Monitor predictions at /admin/
6. 🔄 Retrain to improve accuracy
7. 🚀 Add more features (authentication, history, etc.)

---

Support

Need help?

1. Documentation: See DEPLOYMENT.md
2. Issues: GitHub Issues
3. Discussions: HF Space Discussions

---

Ready? Let's go! 🚀

Choose your deployment method above and follow the steps!