These types represent all the different kinds of sequence that can be used as input of a Tokenizer.
Globally, any sequence can be either a string or a list of strings, according to the operating
mode of the tokenizer: raw text vs pre-tokenized.
tokenizers.TextInputSequence A str that represents an input sequence
tokenizers.PreTokenizedInputSequence A pre-tokenized input sequence. Can be one of:
List of strTuple of stralias of Union[List[str], Tuple[str]].
tokenizers.InputSequence Represents all the possible types of input sequences for encoding. Can be:
is_pretokenized=False: TextInputSequenceis_pretokenized=True: PreTokenizedInputSequencealias of Union[str, List[str], Tuple[str]].