Update app.py

app.py

@@ -677,7 +677,7 @@ def infer_on_image(input_image):
                                        input_width,
                                        input_height)
 
-  return output_image_1, total_time_thinking, output_image_2, total_time_no_thinking
+  return output_image_1, response, total_time_thinking, output_image_2, response_no_thinking, total_time_no_thinking
 
 # Want input to be a single image
 # Outputs to be time + image
@@ -691,8 +691,10 @@ See the app.py file for the different prompts used."""
 demo = gr.Interface(fn=infer_on_image,
                     inputs=gr.Image(label="Input image"),
                     outputs=[gr.Image(label="Image w/ thinking tags"),
+                             gr.Text(label="Raw output w/ thinking tags"),
                              gr.Text(label="Inference time w/ thinking tags"),
                              gr.Image(label="Image w/o thinking tags"),
+                             gr.Text(label="Raw output w/o thinking tags"),
                              gr.Text(label="Inference time w/o thinking tags")],
                     title="Qwen2.5-VL Food Detection 👁️🍔",
                     description=description)