Update app.py

app.py

@@ -241,55 +241,35 @@ def auto_enhance_image_api():
 
     return jsonify({"error": "An unknown error occurred"}), 500
 
-# --- AI IMAGE GENERATOR (TEXT-TO-IMAGE) API ENDPOINT (WITH LAZY LOADING & TINY MODEL) ---
+# --- AI IMAGE GENERATOR (TEXT-TO-IMAGE) API ENDPOINT ---
 
-# Hum model ke liye ek khali "box" (variable) banayenge. Model start mein load nahi hoga.
-pipe = None
+# Model ab pehle se hi download ho chuka hai. Hum usko yahan load karenge.
+# Yeh ab crash nahi hoga kyunki files pehle se cache mein hain.
+pipe = DiffusionPipeline.from_pretrained("hf-internal-testing/tiny-stable-diffusion-torch")
 
 @app.route('/generate-image', methods=['POST'])
 def generate_image_api():
-    global pipe # Hum function ko bata rahe hain ke bahar wale 'pipe' box ko istemal karna hai
-
-    # 1. --- API Key Authentication ---
+    # API Key Authentication
     api_key_header = request.headers.get('x-api-key')
     if not api_key_header or api_key_header != API_KEY:
         return jsonify({"error": "Unauthorized. Invalid or missing API Key."}), 401
 
-    # 2. --- User ka Text Prompt Haasil Karna ---
+    # User ka Text Prompt Haasil Karna
     if not request.is_json:
         return jsonify({"error": "Invalid request: JSON expected"}), 400
-        
     data = request.get_json()
     prompt = data.get('prompt', '')
-
     if not prompt:
         return jsonify({"error": "Prompt is required"}), 400
 
-    # 3. --- AI Model ko Load aur Istemal Karna ---
+    # AI Model se Image Generate Karna
     try:
-        # Check karein ke kya 'pipe' ka box khali hai?
-        if pipe is None:
-            print("Loading Tiny Stable Diffusion model for the first time...")
-            # Use a much smaller and faster model designed for CPUs
-            pipe = DiffusionPipeline.from_pretrained("hf-internal-testing/tiny-stable-diffusion-torch")
-            print("Model loaded successfully.")
-
-        # AI model ko prompt de kar image banana
-        print(f"Generating image for prompt: {prompt}")
-        # Tiny model ke liye kam steps zaroori hain
         image = pipe(prompt, num_inference_steps=10).images[0]
-        print("Image generated.")
-
-        # Ek khali "in-memory" file banana
-        output_buffer = io.BytesIO()
         
-        # Generated image ko buffer mein save karna
+        output_buffer = io.BytesIO()
         image.save(output_buffer, format='PNG')
-        
-        # Image ke bytes haasil karna
         output_image_bytes = output_buffer.getvalue()
 
-        # 4. --- Send the Response ---
         return send_file(
             io.BytesIO(output_image_bytes),
             mimetype='image/png',