Update app.py

app.py

@@ -47,7 +47,7 @@ def get_past_key_values(system_prompt):
     tokenized_test = tokenizer.apply_chat_template(messages, return_tensors='pt')
     assert (tokenized_test[:, :tokenized.shape[1]] == tokenized).all().cpu().item()
     
-    return model(**tokenized.to(model.device)).past_key_values
+    return model(tokenized.to(model.device)).past_key_values
 
 
 @spaces.GPU