Update app.py

app.py

@@ -21,7 +21,7 @@ assistant_prompt = '<|assistant|>\n'
 prompt_suffix = "<|end|>\n"
 
 # Inference function
-def ask_phi(image, question):
+def ask_phi(image, question, max_length):
     start_time = time.time()
     
     prompt = f"{user_prompt}<|image_1|>\n{question}{prompt_suffix}{assistant_prompt}"
@@ -30,7 +30,7 @@ def ask_phi(image, question):
     
     
     params = og.GeneratorParams(model)
-    params.set_search_options(max_length=16384)    
+    params.set_search_options(max_length=max_length)    
     params.set_inputs(inputs)
     generator = og.Generator(model, params)
     
@@ -54,7 +54,8 @@ demo = gr.Interface(
       fn=ask_phi,
       inputs=[
           gr.Image(type="filepath", label="Upload Image"),
-          gr.Textbox(label="Your Prompt")
+          gr.Textbox(label="Your Prompt"),
+          gr.Slider(minimum=16, maximum=16384, step=16, value=4096, label="Tokens")
       ],
       outputs=gr.Textbox(label="Phi-3.5 Response"),
       title="Phi-3.5 Vision Instruct (ONNX)",