Initial commit for Hugging Face deployment

.dockerignore

@@ -0,0 +1,27 @@
+# Python / Backend
+__pycache__/
+*.py[cod]
+*$py.class
+.venv/
+venv/
+env/
+ENV/
+.env
+
+# Large Files / Data
+data/images/
+data/masks/
+*.keras
+*.h5
+*.pkl
+
+# Project
+# frontend/ # Remove this to allow frontend build in multi-stage Dockerfile
+node_modules/
+.git
+.github
+.gitignore
+.dockerignore
+implementation_plan.md
+task.md
+walkthrough.md

Dockerfile

@@ -0,0 +1,39 @@
+# Stage 1: Build the frontend
+FROM node:20-slim AS frontend-builder
+WORKDIR /app/frontend
+COPY frontend/package*.json ./
+RUN npm install
+COPY frontend/ ./
+RUN npm run build
+
+# Stage 2: Final image with Python backend
+FROM python:3.10-slim
+
+WORKDIR /app
+
+# Install system dependencies for OpenCV/Pillow
+RUN apt-get update && apt-get install -y \
+    libglib2.0-0 \
+    libsm6 \
+    libxext6 \
+    libxrender-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy requirements and install
+COPY backend/requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy built frontend from Stage 1
+COPY --from=frontend-builder /app/frontend/dist ./frontend/dist
+
+# Copy backend and supporting code
+COPY backend/ ./backend/
+COPY model/ ./model/
+COPY utils/ ./utils/
+
+# Hugging Face Spaces uses port 7860 by default
+ENV PORT=7860
+EXPOSE 7860
+
+# We use the PORT environment variable in main.py
+CMD ["python", "-m", "uvicorn", "backend.main:app", "--host", "0.0.0.0", "--port", "7860"]

backend/inference.py

@@ -8,6 +8,10 @@ import sys
 sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
 from model.unet import build_unet
 from model.loss_metrics import get_metrics, bce_dice_loss
+from utils.model_fetcher import download_if_missing
+
+# Force download if missing before initialization
+download_if_missing()
 
 # Path to the saved model weights
 MODEL_PATH = os.path.join(os.path.dirname(__file__), "../model/saved_models/oil_spill_unet_best.keras")

backend/main.py

@@ -1,5 +1,6 @@
 from fastapi import FastAPI, UploadFile, File, HTTPException
 from fastapi.responses import Response, JSONResponse
+from fastapi.staticfiles import StaticFiles
 from fastapi.middleware.cors import CORSMiddleware
 from PIL import Image
 import io
@@ -22,6 +23,11 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
+# Serves the built frontend files
+# Ensure 'frontend/dist' exists after the build stage in Docker
+if os.path.exists("frontend/dist"):
+    app.mount("/", StaticFiles(directory="frontend/dist", html=True), name="static")
+
 @app.get("/health")
 def health_check():
     """Health check endpoint to ensure API and Model are ready."""
@@ -69,4 +75,6 @@ async def predict_spill(file: UploadFile = File(...)):
 
 if __name__ == "__main__":
     import uvicorn
-    uvicorn.run("main:app", host="0.0.0.0", port=8000, reload=True)
+    # Hugging Face Spaces defaults to port 7860
+    port = int(os.environ.get("PORT", 7860))
+    uvicorn.run("main:app", host="0.0.0.0", port=port, reload=True)

backend/utils/model_fetcher.py

@@ -0,0 +1,36 @@
+import os
+import urllib.request
+import sys
+
+def download_if_missing():
+    """
+    Checks if the model weights exist locally. 
+    If not, attempts to download them from the URL provided in environment variables.
+    """
+    model_url = os.getenv("MODEL_URL")
+    # Default local path relative to backend directory
+    model_path = os.path.join(os.path.dirname(__file__), "../../model/saved_models/oil_spill_unet_best.keras")
+    
+    # Ensure directory exists
+    os.makedirs(os.path.dirname(model_path), exist_ok=True)
+
+    if not os.path.exists(model_path):
+        if not model_url or model_url == "https://your-r2-public-url/oil_spill_unet_best.keras":
+            print("⚠️ MODEL_URL not configured. Skipping download.")
+            print("💡 Please provide a valid URL in your docker-compose.yml to enable automated fetching.")
+            return
+
+        print(f"🚀 Model weights missing. Initializing download from: {model_url}")
+        try:
+            # Using urllib as it's built-in and sufficient for this simple GET
+            urllib.request.urlretrieve(model_url, model_path)
+            print("✅ Model weights downloaded successfully.")
+        except Exception as e:
+            print(f"❌ Failed to download model: {e}")
+            print("💡 Fallback: The system will use an untrained stub model for now.")
+    else:
+        print("📦 Production model weights detected locally. Skipping download.")
+
+if __name__ == "__main__":
+    # Test run
+    download_if_missing()

docker-compose.yml

@@ -12,6 +12,7 @@ services:
       - ./model:/app/model
     environment:
       - PYTHONUNBUFFERED=1
+      - MODEL_URL=https://pub-b6e6ec55d6e84dcdb100466edde874a2.r2.dev/oil_spill_unet_best.keras
 
   frontend:
     build:

frontend/.dockerignore

@@ -0,0 +1,10 @@
+node_modules
+dist
+build
+.env
+.env.*
+!.env.example
+.git
+.gitignore
+.dockerignore
+README.md

frontend/src/App.tsx

@@ -181,7 +181,7 @@ function Dashboard() {
     formData.append("file", selectedFile);
 
     try {
-      const response = await fetch("http://localhost:8000/predict", {
+      const response = await fetch("/predict", {
         method: "POST",
         body: formData,
       });

upload_to_r2.py

@@ -0,0 +1,41 @@
+import boto3
+import os
+
+# 1. Credentials (Ready to use from your screenshot)
+ACCESS_KEY = "a9f685d13afd9a290e54dca612ef83d2"
+SECRET_KEY = "8284c338275d0e6ebb3e6e697ba5999f6409908cb92c08fe8351da4af728c875"
+ACCOUNT_ID = "659b46b7d1773d3742cd10b2c9194dd3"
+BUCKET_NAME = "oil-spill-detection"
+
+# 2. Path to your model file 
+# Note: Ensure oil_spill_unet_best.keras is in this same folder!
+FILE_PATH = "oil_spill_unet_best.keras" 
+
+def upload():
+    # Setup the Cloudflare R2 Client via S3 API
+    s3 = boto3.client(
+        service_name='s3',
+        endpoint_url=f'https://{ACCOUNT_ID}.r2.cloudflarestorage.com',
+        aws_access_key_id=ACCESS_KEY,
+        aws_secret_access_key=SECRET_KEY,
+        region_name='auto'
+    )
+
+    print(f"Starting high-speed API upload of {FILE_PATH}...")
+    
+    if not os.path.exists(FILE_PATH):
+        print(f"ERROR: File '{FILE_PATH}' not found in the current directory.")
+        print("FIX: Make sure you have downloaded it from Google Drive to this folder.")
+        return
+
+    try:
+        s3.upload_file(FILE_PATH, BUCKET_NAME, "oil_spill_unet_best.keras")
+        print("\n" + "="*40)
+        print("SUCCESS! Your model is now on Cloudflare R2.")
+        print("="*40)
+        print("You can now run 'docker-compose up --build' and the system will pull the model automatically.")
+    except Exception as e:
+        print(f"UPLOAD FAILED: {e}")
+
+if __name__ == "__main__":
+    upload()