Update app.py

app.py

@@ -10,10 +10,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=False)
 
 
 def respond(
-    message,
-    max_tokens,
-    temperature,
-    top_p,
+    message
 ):
     formatted_prompt = f"<s>[INST]{message}[/INST]"
 
@@ -23,9 +20,7 @@ def respond(
     result = model.generate(
         tokens,
         do_sample=False,
-        max_new_tokens=max_tokens,
-        temperature=temperature,
-        top_p=top_p,
+        max_new_tokens=200
     )
 
     response = tokenizer.decode(result[0], skip_special_tokens=True)