Weird warning when calling LlamaTokenizer.__call__

[m-Just]

When I run the following code:

from transformers import LlamaTokenizer

model_path_or_name = "lmsys/vicuna-7b-v1.5"
tokenizer = LlamaTokenizer.from_pretrained(model_path_or_name)

prompt = 'hi'
tokenizer(prompt, return_tensors="pt")

it produces a weird warning that says:
Keyword arguments {'add_special_tokens': False} not recognized.

I traced the warning to this line which calls PreTrainedTokenizer.tokenize with the keyword argument add_special_tokens. It seems that the argument should be handled by self.prepare_for_tokenization inside PreTrainedTokenizer.tokenize, but it is not; leading to the warning.

Transformers version: 4.38.1. (BTW, 4.37.2 does not have this issue)