enabled live drawing and prediction

app.py

@@ -444,9 +444,6 @@ def main():
 
                     image_input =gr.inputs.Image(source="canvas",shape=(28,28),invert_colors=True,image_mode="L",type="pil")
                     label_output = gr.outputs.Label(num_top_classes=2)
-
-
-                submit = gr.Button("Submit")
                 
                 gr.Markdown(MODEL_IS_WRONG)
                 
@@ -459,7 +456,7 @@ def main():
                 adversarial_number = gr.Variable(value=0)
 
 
-                submit.click(image_classifier,inputs = [image_input],outputs=[label_output])
+                image_input.change(image_classifier,inputs = [image_input],outputs=[label_output])
                 flag_btn.click(flag,inputs=[image_input,number_dropdown,adversarial_number],outputs=[output_result,adversarial_number])
                 
             with gr.TabItem('Dashboard') as dashboard:

utils.py

@@ -12,11 +12,10 @@ This kind of data is presumably the most valuable for a model, so this can be he
 """
 WHAT_TO_DO="""
 ### What to do:
-1. Draw any number from 0-9.
-2. Click `Submit` and see the model's prediciton.
-3. If the model misclassifies it, Flag that example.
-4. This will add your (adversarial) example to a dataset on which the model will be trained later.
-5. The model will finetune on the adversarial samples after every __{num_samples}__ samples have been generated. 
+1. Draw any number from 0-9. The model will automatically try to predict it after drawing.
+2. If the model misclassifies it, Flag that example.
+3. This will add your (adversarial) example to a dataset on which the model will be trained later.
+4. The model will finetune on the adversarial samples after every __{num_samples}__ samples have been generated. 
 """
 
 MODEL_IS_WRONG = """