Update app.py

app.py

@@ -19,6 +19,8 @@ processor = AutoProcessor.from_pretrained(model_id)
 model = AutoModelForImageTextToText.from_pretrained(
     model_id, device_map="auto", torch_dtype=torch.bfloat16
 )
+# Update the model's generation config to enable sampling by default
+model.generation_config.do_sample = True
 
 MAX_NUM_IMAGES = int(os.getenv("MAX_NUM_IMAGES", "5"))
 
@@ -188,6 +190,7 @@ def run(message: dict, history: list[dict], system_prompt: str = "", max_new_tok
         inputs,
         max_new_tokens=max_new_tokens,
         streamer=streamer,
+        do_sample=True,
         temperature=1.0,
         top_p=0.95,
         top_k=64,
@@ -217,7 +220,7 @@ demo = gr.ChatInterface(
         gr.Textbox(label="System Prompt", value="You are a helpful medical expert."),
         gr.Slider(label="Max New Tokens", minimum=100, maximum=8192, step=10, value=2048),
     ],
-    stop_btn=False,
+    stop_btn=True,
     title="MedGemma 4B IT",
     description=DESCRIPTION,
     run_examples_on_click=False,