Update app.py

app.py

@@ -27,9 +27,9 @@ llm = LlamaCpp(
 def format_prompt(message, history):
   prompt = "<s>"
   for user_prompt, bot_response in history:
-    prompt += f"[INST] {user_prompt} [/INST]"
-    prompt += f" {bot_response}</s> "
-  prompt += f"[INST] {message} [/INST]"
+    prompt += f"<|im_start|>user\n {user_prompt} <|im_end|>\n"
+    prompt += f"<|im_start|>assistant\n  {bot_response}<|im_end|>\n"
+  prompt += f"<|im_start|>user\n {message} <|im_end|>\n<|im_start|>assistant\n"
   return prompt
 
 def generate(
@@ -51,7 +51,9 @@ def generate(
                temperature=temperature,
         max_tokens=max_new_tokens,
                repeat_penalty=repetition_penalty,
-        top_p=top_p,)
+        top_p=top_p,
+               stop=["<|im_end|>","<|im_start|>user"]
+              )
     # output=formatted_prompt+"ans:"+output
     # for response in stream:
     #     output += response.token.text