Update app.py

app.py

@@ -29,7 +29,7 @@ def generate_image_captions(image, text):
         max_new_tokens=50
     )
 
-    prediction_text = processor.decode(prediction[0], num_beams=5, skip_special_tokens=True)
+    prediction_text = processor.decode(prediction[0], skip_special_tokens=True)
     return prediction_text
 
 
@@ -51,7 +51,7 @@ demo = gr.Interface(
     outputs="text",
     examples=[
         [os.path.join(os.getcwd(), 'two_bear.png'), ""],
-        [os.path.join(os.getcwd(), 'three_women.png'), "What is the woman in the middle's dress's color is? Answer:"],
+        [os.path.join(os.getcwd(), 'three_women.png'), "What is the woman in the middle's dress's color? Answer:"],
         [os.path.join(os.getcwd(), 'cat_with_food.png'), "Describe the picture:"],
         [os.path.join(os.getcwd(), 'dog_with_frisbee.png'), "What is the color of the frisbee in the photo? Answer:"],
         [os.path.join(os.getcwd(), 'stop_sign.png'), "What does the sign in the picture say? Answer:"]