Update app.py

app.py

@@ -43,6 +43,7 @@ model = AutoModelForCausalLM.from_pretrained(
 
 @spaces.GPU()
 def stream_chat(
+    system_prompt: str,
     message: str, 
     history: list, 
     temperature: float = 0.3, 
@@ -54,7 +55,7 @@ def stream_chat(
     print(f'message: {message}')
     print(f'history: {history}')
 
-    conversation = [{"role": "system", "content": "You are EXAONE model from LG AI Research, a helpful assistant."}]
+    conversation = [{"role": "system", "content": system_prompt}]
     for prompt, answer in history:
         conversation.extend([
             {"role": "user", "content": prompt}, 
@@ -105,6 +106,11 @@ with gr.Blocks(css=CSS, theme="soft") as demo:
         fill_height=True,
         additional_inputs_accordion=gr.Accordion(label="⚙️ Parameters", open=False, render=False),
         additional_inputs=[
+            gr.Textbox(
+                value="You are EXAONE model from LG AI Research, a helpful assistant.",
+                label="System Prompt",
+                render=False,
+            ),
             gr.Slider(
                 minimum=0,
                 maximum=1,
@@ -117,7 +123,7 @@ with gr.Blocks(css=CSS, theme="soft") as demo:
                 minimum=128,
                 maximum=4096,
                 step=1,
-                value=256,
+                value=1024,
                 label="Max new tokens",
                 render=False,
             ),