Create app.py

app.py

@@ -0,0 +1,120 @@
+# app.py
+
+import os
+import io
+import base64
+import numpy as np
+from PIL import Image
+import torch
+from realesrgan import RealESRGANer
+
+# FastAPI Libraries
+from fastapi import FastAPI, File, UploadFile
+from fastapi.responses import JSONResponse, StreamingResponse
+import uvicorn
+import gradio as gr
+
+# --- 1. Model Loading (Free Tier Optimized) ---
+
+DEVICE = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+print(f"Model will run on: {DEVICE}")
+
+try:
+    # Real-ESRGAN ka lightweight, optimized model use kar rahe hain
+    model_path = RealESRGANer.model_path_from_name('RealESRGAN_x4plus')
+    
+    # Model ko load karna (yeh memory mein rahega)
+    UPSCALER = RealESRGANer(
+        scale=4, 
+        model_path=model_path,
+        dni_weight=None,
+        model_name='RealESRGAN_x4plus',
+        device=DEVICE
+    )
+    print("Real-ESRGAN model loaded successfully.")
+
+except Exception as e:
+    print(f"ERROR: Model load nahi ho paya. Error: {e}")
+    UPSCALER = None
+
+def run_upscaler(img_np: np.ndarray):
+    """Core upscaling logic."""
+    if UPSCALER is None:
+        raise Exception("Model is not initialized.")
+        
+    # Upscaling (yahan time lagta hai)
+    output_np, _ = UPSCALER.enhance(img_np, outscale=4)
+    
+    return output_np
+
+# --- 2. FastAPI Setup ---
+
+# FastAPI application ko initialize karein
+app = FastAPI(title="Real-ESRGAN Custom Upscaler API")
+
+# --- 3. Custom API Endpoint ---
+
+# Image file upload ke zariye upscaling
+@app.post("/api/upscale/file")
+async def upscale_image_api(image: UploadFile = File(...)):
+    """
+    Image file ko upload karein aur 4x upscaled image wapas hasil karein.
+    """
+    try:
+        # File ko PIL Image mein padhna
+        image_bytes = await image.read()
+        input_image = Image.open(io.BytesIO(image_bytes)).convert("RGB")
+        
+        # PIL image ko numpy array mein convert karna
+        img_np = np.array(input_image)
+        
+        # Upscaling
+        output_np = run_upscaler(img_np)
+        
+        # NumPy array ko wapas PIL Image mein convert karna
+        output_image = Image.fromarray(output_np)
+        
+        # Image ko BytesIO mein save karna
+        img_io = io.BytesIO()
+        output_image.save(img_io, format='PNG')
+        img_io.seek(0)
+        
+        # StreamingResponse se image ko wapas bhejna
+        return StreamingResponse(img_io, media_type="image/png")
+
+    except Exception as e:
+        return JSONResponse(status_code=500, content={"message": f"Processing error: {str(e)}"})
+
+# --- 4. Gradio UI Integration ---
+
+def upscale_for_gradio(input_image: Image.Image):
+    """Gradio UI ke liye wrapper function."""
+    try:
+        img_np = np.array(input_image.convert("RGB"))
+        output_np = run_upscaler(img_np)
+        return Image.fromarray(output_np)
+    except Exception as e:
+        return f"Error: {str(e)}"
+
+# Gradio Interface define karna
+gr_interface = gr.Interface(
+    fn=upscale_for_gradio,
+    inputs=gr.Image(type="pil", label="Low-Resolution Image Upload Karein"),
+    outputs=gr.Image(type="pil", label="4x Upscaled (High-Quality) Image"),
+    title="⭐ Real-ESRGAN: AI Image Upscaler (UI & Custom API)",
+    description="Apni images ko 4x size mein badhayein. Yeh app Custom REST API aur Gradio UI dono offer karta hai.",
+    allow_flagging="never"
+)
+
+# Gradio ko FastAPI app mein mount karna
+# '/gradio' path par UI available hoga
+app = gr.mount_gradio_app(app, gr_interface, path="/") 
+
+# --- 5. Uvicorn Server Setup ---
+
+# Yeh tabhi run hoga jab aap file ko directly chalayenge (lekin Docker mein yeh entry point hoga)
+if __name__ == "__main__":
+    # Hugging Face Spaces Docker mein port 7860 par chalne ki umeed rakhta hai.
+    # Hamara server isi port par run hoga.
+    uvicorn.run(app, host="0.0.0.0", port=7860)
+