Update app.py

app.py

@@ -33,8 +33,7 @@ def respond(
     temperature,
     top_p,
     peft_model,
-    tokenizer_name='your-tokenizer-name',
-    device='auto'  # or 'cpu' based on your setup
+    tokenizer_name,
 ):
     """
     Generates a response based on the user message and history using the provided PEFT model.
@@ -48,7 +47,6 @@ def respond(
         top_p (float): The top_p parameter for nucleus sampling.
         peft_model: The pre-trained fine-tuned model for generation.
         tokenizer_name (str): The name or path of the tokenizer.
-        device (str): The device to run the model on ('cuda' or 'cpu').
 
     Yields:
         str: The generated response up to the current token.