Description

inputs = processor.apply_chat_template(
    messages,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
    enable_thinking=False,
    #chat_template_kwargs={"enable_thinking": False}
).to(qwen_model.device)  

I use this code to run qwen3.5_9b, but the resulit is "enable_thinking=False is not a valid argument" , How can I close the thinking mode?

Reproduction

inputs = processor.apply_chat_template(
messages,
tokenize=True,
return_dict=True,
return_tensors="pt",
enable_thinking=False,
#chat_template_kwargs={"enable_thinking": False}
).to(qwen_model.device)

Logs

Environment Information

transformers ==5.3

Known Issue

• The issue hasn't been already addressed in Documentation, Issues, and Discussions.