Update app.py

app.py

@@ -88,6 +88,7 @@ title = "From Images to Textual Prompts: Zero-shot VQA with Frozen Large Languag
 df_init = pd.DataFrame(columns=['Caption'])
 raw_image = gr.Image(label='Input image', type="pil")
 question = gr.Textbox(label="Input question", lines=1, interactive=True)
+text_output = gr.Textbox(label="Output Answer")
 demo = gr.Blocks(title=title)
 demo.encrypt = False
 cap_df = gr.DataFrame(value=df_init, label="Caption dataframe", row_count=(0, "dynamic"), max_rows = 20, wrap=True, overflow_row_behaviour='paginate')
@@ -113,29 +114,24 @@ with demo:
     label="Examples", inputs=[raw_image, question])
     
     with gr.Row():
-      with gr.Column():
-          raw_image.render()
-      with gr.Column():
-          avg_gradcam = gr.Image(label="GradCam image",)
-
+        with gr.Column():  
+            raw_image.render()
+        with gr.Column():
+            question.render()
+            number_cap = gr.Number(precision=0, value=5, label="Selected number of caption you want to generate", interactive=True)
     with gr.Row():
       with gr.Column():
-          question.render()
+            cap_btn = gr.Button("Generate caption")
+            cap_btn.click(generate_cap, [raw_image, question, number_cap], [cap_df])
       with gr.Column():
-          number_cap = gr.Number(precision=0, value=5, label="Selected number of caption you want to generate", interactive=True)
+            anws_btn = gr.Button("Answer")
+            anws_btn.click(generate_answer, [raw_image, question], outputs=text_output)
     with gr.Row():  
       with gr.Column():  
-          gradcam_btn = gr.Button("Generate Gradcam")
-          gradcam_btn.click(gradcam_attention, [raw_image, question], outputs=[avg_gradcam])
-      with gr.Column():
-          cap_btn = gr.Button("Generate caption")
-          cap_btn.click(generate_cap, [raw_image, question, number_cap], [cap_df])
-    with gr.Row():
-      with gr.Column():
-          cap_df.render()
-    with gr.Row():
-        anws_btn = gr.Button("Answer")
-        text_output = gr.Textbox()
-        anws_btn.click(generate_answer, [raw_image, question], outputs=text_output)
+      #     gradcam_btn = gr.Button("Generate Gradcam")
+      #     gradcam_btn.click(gradcam_attention, [raw_image, question], outputs=[avg_gradcam])
+            cap_df.render()
+      with gr.Column():  
+            text_output.render()
 
 demo.launch(debug=True)