feat: Final working version with FastAPI deployment

README.md

@@ -64,17 +64,18 @@ The project follows a structured MLOps workflow, which is fully automated by the
 
 ```mermaid
 graph TD
-    A["Start: Push Code to GitHub"] --> B{"GitHub Actions CI/CD"};
-    B --> C["CI: Install Dependencies & Run Tests"];
-    C -->|Success| D["CD: Deploy to Hugging Face"];
-    D --> E["🚀 Live Application"];
+    A[Start: Push Code to GitHub] --> B{GitHub Actions CI/CD};
+    B --> C[CI: Install Dependencies & Run Tests];
+    C -->|Success| D[CD: Deploy to Hugging Face];
+    D --> E[🚀 Live Application];
 
     subgraph "DVC & MLflow Cycle (Local/Remote)"
-        F["1. `dvc repro`"] --> G["2. Pull Data (DVC)"];
-        G --> H["3. Train Model"];
-        H --> I["4. Log Metrics & Model (MLflow)"];
-        I --> J["5. Push Model (DVC)"];
+        F[1. `dvc repro`] --> G[2. Pull Data (DVC)];
+        G --> H[3. Train Model];
+        H --> I[4. Log Metrics & Model (MLflow)];
+        I --> J[5. Push Model (DVC)];
     end
+```
 
 ## ⚙️ Getting Started - Local Setup
 

app.py

@@ -1,55 +1,98 @@
-from flask import Flask, request, jsonify, render_template
+# --- IMPORTS ---
 import os
-from flask_cors import CORS, cross_origin
+from fastapi import FastAPI, Request
+from fastapi.responses import HTMLResponse
+from fastapi.staticfiles import StaticFiles
+from fastapi.templating import Jinja2Templates
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+import uvicorn # For running the app directly if needed
+
+# Your existing ML components
 from cnnClassifier.utils.common import decodeImage
 from cnnClassifier.pipeline.prediction import PredictionPipeline
 
-# Set environment variables for consistent encoding
+# --- CONFIGURATION ---
 os.putenv('LANG', 'en_US.UTF-8')
 os.putenv('LC_ALL', 'en_US.UTF-8')
 
-app = Flask(__name__)
-CORS(app)
-
-class ClientApp:
-    def __init__(self):
-        self.filename = "inputImage.jpg"
-        self.classifier = PredictionPipeline(self.filename)
-
-@app.route("/", methods=['GET'])
-@cross_origin()
-def home():
-    """Renders the main user interface."""
-    return render_template('index.html')
-
-@app.route("/train", methods=['GET','POST'])
-@cross_origin()
-def trainRoute():
-    """Triggers the DVC pipeline to retrain the model."""
-    # os.system("python main.py") # You can use this if you have a main orchestrator
+# --- INITIALIZE FastAPI APP ---
+app = FastAPI(
+    title="Chest Cancer Classification API",
+    description="An API to predict whether a chest CT scan shows signs of adenocarcinoma cancer."
+)
+
+# --- MIDDLEWARE (for CORS) ---
+# This is the FastAPI equivalent of Flask-CORS
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+# --- MOUNT STATIC FILES AND TEMPLATES ---
+# This is how FastAPI serves your CSS, JS, and HTML files
+app.mount("/static", StaticFiles(directory="static"), name="static")
+templates = Jinja2Templates(directory="templates")
+
+
+# --- LOAD THE PREDICTION PIPELINE ON STARTUP ---
+# This ensures the model is loaded only once when the application starts.
+classifier = PredictionPipeline(filename="inputImage.jpg")
+
+
+# --- DEFINE THE REQUEST BODY STRUCTURE ---
+# Pydantic model for automatic validation of the incoming JSON
+class ImagePayload(BaseModel):
+    image: str
+
+
+# --- API ENDPOINTS ---
+
+@app.get("/", response_class=HTMLResponse)
+async def home(request: Request):
+    """
+    Renders the main user interface (index.html).
+    """
+    return templates.TemplateResponse("index.html", {"request": request})
+
+
+@app.post("/train")
+async def trainRoute():
+    """
+    Triggers the DVC pipeline to retrain the model.
+    NOTE: This is a blocking operation and not recommended for a real-world, high-traffic production server.
+    It's suitable for this project's demonstration purposes.
+    """
     os.system("dvc repro")
-    return "Training done successfully!"
-
-@app.route("/predict", methods=['POST'])
-@cross_origin()
-def predictRoute():
-    image = request.json['image']
-    decodeImage(image, clApp.filename)
-    
-    # The predict() method now returns just the index (0 or 1)
-    prediction_value = clApp.classifier.predict()
-    
-    # This logic is confirmed by your class indices: {'adenocarcinoma': 0, 'normal': 1}
+    return {"message": "Training done successfully!"}
+
+
+@app.post("/predict")
+async def predictRoute(payload: ImagePayload):
+    """
+    Accepts a base64 encoded image, saves it, runs prediction, and returns the result.
+    """
+    # 1. Decode the image and save it
+    decodeImage(payload.image, "inputImage.jpg")
+
+    # 2. Run the prediction pipeline
+    prediction_value = classifier.predict()
+
+    # 3. Translate the numeric prediction into a human-readable string
+    # Based on your confirmed class indices: {'adenocarcinoma': 0, 'normal': 1}
     if prediction_value == 1:
         prediction_text = "Normal"
-    else: # The value was 0
+    else:  # The value was 0
         prediction_text = "Cancer"
 
-    # The front-end expects the key "prediction"
-    return jsonify([{"prediction": prediction_text}])
+    # 4. Return the result. FastAPI handles the JSON conversion.
+    return [{"prediction": prediction_text}]
 
 
+# --- RUN THE APP ---
+# This block is for local development. Gunicorn/Uvicorn will run the app in production.
 if __name__ == "__main__":
-    clApp = ClientApp()
-    # Run the app on all available interfaces (for Docker/deployment) and port 8080
-    app.run(host='0.0.0.0', port=8080)
+    uvicorn.run(app, host="0.0.0.0", port=8001)

dockerfile

@@ -1,13 +1,7 @@
 FROM python:3.8-slim
-
 WORKDIR /app
-
 COPY requirements.txt .
-
 RUN pip install --no-cache-dir -r requirements.txt
-
 COPY . .
-
 EXPOSE 8080
-
-CMD ["gunicorn", "--bind", "0.0.0.0:8080", "app:app"]
+CMD ["    pip install uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8080"]

templates/index.html

@@ -17,7 +17,7 @@
     <link href="https://fonts.googleapis.com/css2?family=Poppins:wght@300;400;500;600;700&display=swap" rel="stylesheet">
     
     <!-- Your Custom CSS -->
-    <link rel="stylesheet" href="{{ url_for('static', filename='style.css') }}">
+    <link rel="stylesheet" href="{{ url_for('static', path='style.css') }}">
 </head>
 <body>
 
@@ -93,6 +93,6 @@
     <!-- Bootstrap 5 JS -->
     <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.3.2/dist/js/bootstrap.bundle.min.js"></script>
     <!-- Your Custom JS -->
-    <script src="{{ url_for('static', filename='script.js') }}"></script>
+    <script src="{{ url_for('static', path='script.js') }}"></script>
 </body>
 </html>