Update app.py

app.py

@@ -192,7 +192,7 @@ def inference(raw_image, question):
   mask = cv2.resize(mask / mask.max(), oimg.size)[..., np.newaxis]
   attention_map = (mask * oimg).astype("uint8")
   splitted_seq = seq.split("because")
-  return splitted_seq[0].strip(), "because " + " ".join(splitted_seq[-1].strip()), Image.fromarray(attention_map)
+  return splitted_seq[0].strip(), "because " + splitted_seq[-1].strip(), Image.fromarray(attention_map)
 
 inputs = [gr.inputs.Image(type='pil', label="Load the image of your interest"), gr.inputs.Textbox(label="Ask a question on this image")]
 outputs = [gr.outputs.Textbox(label="Answer"), gr.outputs.Textbox(label="Textual Explanation"), gr.outputs.Image(type='pil', label="Visual Explanation")]