Update app.py

app.py

@@ -3,6 +3,7 @@ from transformers import ViltProcessor, ViltForVisualQuestionAnswering
 import torch
 
 torch.hub.download_url_to_file('http://images.cocodataset.org/val2017/000000039769.jpg', 'cats.jpg')
+torch.hub.download_url_to_file('https://computing.ece.vt.edu/~harsh/visualAttention/ProjectWebpage/Figures/vqa_1.png', 'banana.png')
 
 processor = ViltProcessor.from_pretrained("dandelin/vilt-b32-finetuned-vqa")
 model = ViltForVisualQuestionAnswering.from_pretrained("dandelin/vilt-b32-finetuned-vqa")
@@ -23,24 +24,7 @@ def answer_question(image, text):
 image = gr.inputs.Image(type="pil")
 question = gr.inputs.Textbox(label="Question")
 answer = gr.outputs.Textbox(label="Predicted answer")
-examples = [["cats.jpg", "How many cats are there?"], 
-            [
-                "https://s3.geograph.org.uk/geophotos/06/21/24/6212487_1cca7f3f_1024x1024.jpg",
-                "What is the color of the flower?",
-            ],
-            [
-                "https://computing.ece.vt.edu/~harsh/visualAttention/ProjectWebpage/Figures/vqa_1.png",
-                "What is the mustache made of?",
-            ],
-            [
-                "https://computing.ece.vt.edu/~harsh/visualAttention/ProjectWebpage/Figures/vqa_2.png",
-                "How many slices of pizza are there?",
-            ],
-            [
-                "https://computing.ece.vt.edu/~harsh/visualAttention/ProjectWebpage/Figures/vqa_3.png",
-                "Does it appear to be rainy?",
-            ],
-]
+examples = [["cats.jpg", "How many cats are there?"], ["banana.png", "What is the mustache made of?"]]
 
 interface = gr.Interface(fn=answer_question, inputs=[image, question], outputs=answer, examples=examples, enable_queue=True)
 interface.launch(debug=True)