Update app.py

app.py

@@ -5,9 +5,10 @@ client = InferenceClient(
     "mistralai/Mistral-7B-Instruct-v0.1"
 )
 
-
-def format_prompt(message, history):
+def format_prompt(message, history, system_prompt=""):
   prompt = "<s>"
+  if system_prompt:
+      prompt += f"[SYS] {system_prompt} [/SYS] "
   for user_prompt, bot_response in history:
     prompt += f"[INST] {user_prompt} [/INST]"
     prompt += f" {bot_response}</s> "
@@ -15,7 +16,7 @@ def format_prompt(message, history):
   return prompt
 
 def generate(
-    prompt, history, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
+    prompt, history, system_prompt="", temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
 ):
     temperature = float(temperature)
     if temperature < 1e-2:
@@ -31,7 +32,7 @@ def generate(
         seed=42,
     )
 
-    formatted_prompt = format_prompt(prompt, history)
+    formatted_prompt = format_prompt(prompt, history, system_prompt)
 
     stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
     output = ""
@@ -41,8 +42,8 @@ def generate(
         yield output
     return output
 
-
 additional_inputs=[
+    gr.Textbox("", label="Optional system prompt"),
     gr.Slider(
         label="Temperature",
         value=0.9,
@@ -99,4 +100,4 @@ with gr.Blocks(css=css) as demo:
         examples=[["What is the secret to life?"], ["Write me a recipe for pancakes."]]
     )
 
-demo.queue().launch(debug=True)
+demo.queue().launch(debug=True)