fix

app.py

@@ -25,6 +25,12 @@ MAX_TOKENS = 4096
 print(f"[Init] Loading model pipeline for the first time: {MODEL_NAME}...")
 @spaces.GPU
 def load_model():
+    
+
+
+@spaces.GPU() 
+@torch.inference_mode()
+def predict(image: Image.Image) -> str:
     try:
         from unsloth import FastVisionModel
         model, processor = FastVisionModel.from_pretrained(
@@ -34,18 +40,12 @@ def load_model():
                     use_gradient_checkpointing="unsloth",
                     token=os.environ["HF_API_KEY"],
                 )
-        return model, processor
     except Exception as e:
         print(f"[Error] Failed to load model: {e}")
         raise  Exception(f"❌ Model failed to load: {e}")
-
-model, processor = load_model()
-
-@spaces.GPU() 
-@torch.inference_mode()
-def predict(image: Image.Image) -> str:
+        
     if image is None:
-        return "Please upload an image."
+        gr.warning("Please upload an image.")
 
     # Build prompt
     messages = [