docs/DEPLOYMENT_GUIDE.md

# Deployment Guide - Two Options

You now have TWO deployment options for your model:

## Option 1: Gradio Space (Current - Interactive UI)

**URL**: https://huggingface.co/spaces/chenhaoq87/MilkSpoilageClassifier-Demo

**Features**:
- ✅ Interactive web interface
- ✅ Visual inputs and outputs
- ❌ No direct REST API for Custom GPT

**Best for**: Human users testing the model manually

---

## Option 2: FastAPI Space (Recommended for Custom GPT)

Create a **second Space** with pure REST API:

### Step 1: Create New Space

1. Go to https://huggingface.co/new-space
2. Name: `MilkSpoilageClassifier-API`
3. SDK: **Docker**
4. Click "Create Space"

### Step 2: Upload Files

Upload these files to the new Space:

**Files to upload:**
- `apps/fastapi/app.py` → upload as `app.py`
- `model/model.joblib` → upload as `model.joblib`
- `apps/fastapi/Dockerfile` → upload as `Dockerfile`
- `apps/fastapi/requirements.txt` → upload as `requirements.txt`

**Dockerfile:**
```dockerfile
FROM python:3.10-slim

WORKDIR /app

# Copy files
COPY requirements.txt .
COPY app.py .
COPY model.joblib .

# Install dependencies
RUN pip install --no-cache-dir -r requirements.txt

# Expose port
EXPOSE 7860

# Run app
CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]
```

**requirements.txt**:
```
fastapi>=0.100
uvicorn[standard]>=0.24
pydantic>=2.0
scikit-learn>=1.0
joblib>=1.0
numpy>=1.20
```

### Step 3: Use with Custom GPT

Once the Space is running, your REST API will be at:
```
https://chenhaoq87-milkspoilageclassifier-api.hf.space/predict
```

**OpenAPI Schema** (see API_DOCUMENTATION.md)

---

## Quick Command to Create FastAPI Space

Run this in PowerShell:

```powershell
cd D:\HuggingFace\MilkSpoilageClassifier

# Upload to new Space
python -c "from huggingface_hub import HfApi, create_repo; api = HfApi(); create_repo('chenhaoq87/MilkSpoilageClassifier-API', repo_type='space', space_sdk='docker', exist_ok=True); api.upload_file('apps/fastapi/Dockerfile', 'Dockerfile', 'chenhaoq87/MilkSpoilageClassifier-API', repo_type='space'); api.upload_file('apps/fastapi/app.py', 'app.py', 'chenhaoq87/MilkSpoilageClassifier-API', repo_type='space'); api.upload_file('apps/fastapi/requirements.txt', 'requirements.txt', 'chenhaoq87/MilkSpoilageClassifier-API', repo_type='space'); api.upload_file('model/model.joblib', 'model.joblib', 'chenhaoq87/MilkSpoilageClassifier-API', repo_type='space'); print('FastAPI Space created!')"
```

---

## Summary

| Feature | Gradio Space | FastAPI Space |
|---------|-------------|---------------|
| Interactive UI | ✅ | ❌ |
| REST API | ❌ | ✅ |
| Custom GPT | ❌ | ✅ |
| Human Testing | ✅ | ❌ |
| API Docs | ❌ | ✅ (/docs) |

**Recommendation**: Keep both!
- Use Gradio Space for manual testing
- Use FastAPI Space for Custom GPT integration