Update app.py

app.py

@@ -61,7 +61,7 @@ def generate(
     top_k,
     top_p,
     max_new_tokens,
-    repetition_penalty
+    repetition_penalty,
 ):
     client = Client(
         model2endpoint[model_name]
@@ -109,7 +109,7 @@ def generate(
         top_p=top_p,
         repetition_penalty=repetition_penalty,
         do_sample=True,
-        truncate=999,
+        truncate=1000,
         seed=42,
         stop_sequences=["<|end|>"],
     )
@@ -138,7 +138,6 @@ def generate(
 
         yield chat, history, user_message, ""
 
-
     return chat, history, user_message, ""
 
 
@@ -164,7 +163,7 @@ def process_example(args):
     return [x, y]
 
 
-title = """<h1 align="center">⭐ StarChat Saturdays 💬</h1>"""
+title = """<h1 align="center">⭐ StarChat Playground 💬</h1>"""
 custom_css = """
 #banner-image {
     display: block;
@@ -181,7 +180,8 @@ custom_css = """
 with gr.Blocks(analytics_enabled=False, css=custom_css) as demo:
     gr.HTML(title)
 
-    selected_model = model_names[0]
+    with gr.Row():
+        selected_model = gr.Radio(choices=model_names, value=model_names[0], label="Select a model")
 
     with gr.Accordion(label="System Prompt", open=False, elem_id="parameters-accordion"):
         system_message = gr.Textbox(
@@ -234,9 +234,9 @@ with gr.Blocks(analytics_enabled=False, css=custom_css) as demo:
                 )
                 max_new_tokens = gr.Slider(
                     label="Max new tokens",
-                    value=512,
+                    value=1024,
                     minimum=0,
-                    maximum=1024,
+                    maximum=2048,
                     step=4,
                     interactive=True,
                     info="The maximum numbers of new tokens",
@@ -305,4 +305,4 @@ with gr.Blocks(analytics_enabled=False, css=custom_css) as demo:
     selected_model.change(clear_chat, outputs=[chatbot, history])
     # share_button.click(None, [], [], _js=share_js)
 
-demo.queue(concurrency_count=16).launch(debug=True)
+demo.queue(concurrency_count=16).launch()