tokenization_utils_base

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase

Bases: SpecialTokensMixin

Base class for [PreTrainedTokenizer] and [PreTrainedTokenizerFast].

Handles shared (mostly boiler plate) methods for those two classes.

Source code in mindnlp\transformers\tokenization_utils_base.py

class PreTrainedTokenizerBase(SpecialTokensMixin):
    """
    Base class for [`PreTrainedTokenizer`] and [`PreTrainedTokenizerFast`].

    Handles shared (mostly boiler plate) methods for those two classes.
    """

    vocab_files_names: Dict[str, str] = {}
    pretrained_vocab_files_map: Dict[str, Dict[str, str]] = {}
    _auto_class: Optional[str] = None

    # first name has to correspond to main model input name
    # to make sure `tokenizer.pad(...)` works correctly
    model_input_names: List[str] = ["input_ids", "token_type_ids", "attention_mask"]
    padding_side: str = "right"
    truncation_side: str = "right"
    slow_tokenizer_class = None

    def __init__(self, **kwargs):
        # inputs and kwargs for saving and re-loading (see ``from_pretrained`` and ``save_pretrained``)
        self.init_inputs = ()
        for key in kwargs:
            if hasattr(self, key) and callable(getattr(self, key)):
                raise AttributeError(f"{key} conflicts with the method {key} in {self.__class__.__name__}")

        self.init_kwargs = copy.deepcopy(kwargs)
        self.name_or_path = kwargs.pop("name_or_path", "")
        self._processor_class = kwargs.pop("processor_class", None)

        # For backward compatibility we fallback to set model_max_length from max_len if provided
        model_max_length = kwargs.pop("model_max_length", kwargs.pop("max_len", None))
        self.model_max_length = model_max_length if model_max_length is not None else VERY_LARGE_INTEGER

        # Padding and truncation side are right by default and overridden in subclasses. If specified in the kwargs, it
        # is changed.
        self.padding_side = kwargs.pop("padding_side", self.padding_side)
        if self.padding_side not in ["right", "left"]:
            raise ValueError(
                f"Padding side should be selected between 'right' and 'left', current value: {self.padding_side}"
            )

        self.truncation_side = kwargs.pop("truncation_side", self.truncation_side)
        if self.truncation_side not in ["right", "left"]:
            raise ValueError(
                f"Truncation side should be selected between 'right' and 'left', current value: {self.truncation_side}"
            )

        self.model_input_names = kwargs.pop("model_input_names", self.model_input_names)

        if "clean_up_tokenization_spaces" not in kwargs:
            warnings.warn(
                "`clean_up_tokenization_spaces` was not set. It will be set to `True` by default. This "
                "behavior will be depracted, and will be then set to `False` by default. ",
                FutureWarning,
            )

        # By default, cleaning tokenization spaces for both fast and slow tokenizers
        self.clean_up_tokenization_spaces = kwargs.pop("clean_up_tokenization_spaces", True)

        # By default, do not split special tokens for both fast and slow tokenizers
        self.split_special_tokens = kwargs.pop("split_special_tokens", False)

        self.deprecation_warnings = {}  # Use to store when we have already noticed a deprecation warning (avoid overlogging).
        self._in_target_context_manager = False

        # Stores a Jinja template that formats chat histories into tokenizable strings
        self.chat_template = kwargs.pop("chat_template", None)
        if isinstance(self.chat_template, (list, tuple)):
            # Chat templates are stored as lists of dicts with fixed key names,
            # we reconstruct that into a single dict while loading them.
            self.chat_template = {template["name"]: template["template"] for template in self.chat_template}

        super().__init__(**kwargs)

    @property
    def max_len_single_sentence(self) -> int:
        """
        `int`: The maximum length of a sentence that can be fed to the model.
        """
        return self.model_max_length - self.num_special_tokens_to_add(pair=False)

    @property
    def max_len_sentences_pair(self) -> int:
        """
        `int`: The maximum combined length of a pair of sentences that can be fed to the model.
        """
        return self.model_max_length - self.num_special_tokens_to_add(pair=True)

    @max_len_single_sentence.setter
    def max_len_single_sentence(self, value) -> int:
        # For backward compatibility, allow to try to setup 'max_len_single_sentence'.
        if value == self.model_max_length - self.num_special_tokens_to_add(pair=False) and self.verbose:
            if not self.deprecation_warnings.get("max_len_single_sentence", False):
                logger.warning(
                    "Setting 'max_len_single_sentence' is now deprecated. This value is automatically set up."
                )
            self.deprecation_warnings["max_len_single_sentence"] = True
        else:
            raise ValueError(
                "Setting 'max_len_single_sentence' is now deprecated. This value is automatically set up."
            )

    @max_len_sentences_pair.setter
    def max_len_sentences_pair(self, value) -> int:
        # For backward compatibility, allow to try to setup 'max_len_sentences_pair'.
        if value == self.model_max_length - self.num_special_tokens_to_add(pair=True) and self.verbose:
            if not self.deprecation_warnings.get("max_len_sentences_pair", False):
                logger.warning(
                    "Setting 'max_len_sentences_pair' is now deprecated. This value is automatically set up."
                )
            self.deprecation_warnings["max_len_sentences_pair"] = True
        else:
            raise ValueError("Setting 'max_len_sentences_pair' is now deprecated. This value is automatically set up.")

    def _set_processor_class(self, processor_class: str):
        """Sets processor class as an attribute."""
        self._processor_class = processor_class

    @property
    def added_tokens_decoder(self) -> Dict[int, AddedToken]:
        raise NotImplementedError()

    def __repr__(self) -> str:
        added_tokens_decoder_rep = "\n\t".join([f"{k}: {v.__repr__()}," for k, v in self.added_tokens_decoder.items()])
        return (
            f"{self.__class__.__name__}(name_or_path='{self.name_or_path}',"
            f" vocab_size={self.vocab_size}, model_max_length={self.model_max_length}, is_fast={self.is_fast},"
            f" padding_side='{self.padding_side}', truncation_side='{self.truncation_side}',"
            f" special_tokens={self.special_tokens_map}, clean_up_tokenization_spaces={self.clean_up_tokenization_spaces}), "
            " added_tokens_decoder={\n\t" + added_tokens_decoder_rep + "\n}"
        )

    def __len__(self) -> int:
        raise NotImplementedError()

    def get_vocab(self) -> Dict[str, int]:
        """
        Returns the vocabulary as a dictionary of token to index.

        `tokenizer.get_vocab()[token]` is equivalent to `tokenizer.convert_tokens_to_ids(token)` when `token` is in the
        vocab.

        Returns:
            `Dict[str, int]`: The vocabulary.
        """
        raise NotImplementedError()

    def apply_chat_template(
        self,
        conversation: Union[List[Dict[str, str]], List[List[Dict[str, str]]]],
        tools: Optional[List[Dict]] = None,
        documents: Optional[List[Dict[str, str]]] = None,
        chat_template: Optional[str] = None,
        add_generation_prompt: bool = False,
        tokenize: bool = True,
        padding: bool = False,
        truncation: bool = False,
        max_length: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_dict: bool = False,
        return_assistant_tokens_mask: bool = False,
        tokenizer_kwargs: Optional[Dict[str, Any]] = None,
        **kwargs,
    ) -> Union[str, List[int], List[str], List[List[int]], BatchEncoding]:
        """
        Converts a list of dictionaries with `"role"` and `"content"` keys to a list of token
        ids. This method is intended for use with chat models, and will read the tokenizer's chat_template attribute to
        determine the format and control tokens to use when converting.

        Args:
            conversation (Union[List[Dict[str, str]], List[List[Dict[str, str]]]]): A list of dicts
                with "role" and "content" keys, representing the chat history so far.
            tools (`List[Dict]`, *optional*):
                A list of tools (callable functions) that will be accessible to the model. If the template does not
                support function calling, this argument will have no effect. Each tool should be passed as a JSON Schema,
                giving the name, description and argument types for the tool. See our
                [chat templating guide](https://huggingface.co/docs/transformers/main/en/chat_templating#automated-function-conversion-for-tool-use)
                for more information.
            documents (`List[Dict[str, str]]`, *optional*):
                A list of dicts representing documents that will be accessible to the model if it is performing RAG
                (retrieval-augmented generation). If the template does not support RAG, this argument will have no
                effect. We recommend that each document should be a dict containing "title" and "text" keys. Please
                see the RAG section of the [chat templating guide](https://huggingface.co/docs/transformers/main/en/chat_templating#arguments-for-RAG)
                for examples of passing documents with chat templates.
            chat_template (`str`, *optional*):
                A Jinja template to use for this conversion. It is usually not necessary to pass anything to this
                argument, as the model's template will be used by default.
            add_generation_prompt (bool, *optional*): Whether to end the prompt with the token(s) that indicate
                the start of an assistant message. This is useful when you want to generate a response from the model.
                Note that this argument will be passed to the chat template, and so it must be supported in the
                template for this argument to have any effect.
            tokenize (`bool`, defaults to `True`):
                Whether to tokenize the output. If `False`, the output will be a string.
            padding (`bool`, defaults to `False`):
                Whether to pad sequences to the maximum length. Has no effect if tokenize is `False`.
            truncation (`bool`, defaults to `False`):
                Whether to truncate sequences at the maximum length. Has no effect if tokenize is `False`.
            max_length (`int`, *optional*):
                Maximum length (in tokens) to use for padding or truncation. Has no effect if tokenize is `False`. If
                not specified, the tokenizer's `max_length` attribute will be used as a default.
            return_tensors (`str` or [`~utils.TensorType`], *optional*):
                If set, will return tensors of a particular framework. Has no effect if tokenize is `False`. Acceptable
                values are:
                - `'tf'`: Return TensorFlow `tf.Tensor` objects.
                - `'pt'`: Return PyTorch `mindspore.Tensor` objects.
                - `'np'`: Return NumPy `np.ndarray` objects.
                - `'jax'`: Return JAX `jnp.ndarray` objects.
            return_dict (`bool`, defaults to `False`):
                Whether to return a dictionary with named outputs. Has no effect if tokenize is `False`.
            tokenizer_kwargs (`Dict[str: Any]`, *optional*): Additional kwargs to pass to the tokenizer.
            return_assistant_tokens_mask (`bool`, defaults to `False`):
                Whether to return a mask of the assistant generated tokens. For tokens generated by the assistant,
                the mask will contain 1. For user and system tokens, the mask will contain 0.
                This functionality is only available for chat templates that support it via the `{% generation %}` keyword.
            **kwargs: Additional kwargs to pass to the template renderer. Will be accessible by the chat template.

        Returns:
            `Union[List[int], Dict]`: A list of token ids representing the tokenized chat so far, including control tokens. This
            output is ready to pass to the model, either directly or via methods like `generate()`. If `return_dict` is
            set, will return a dict of tokenizer outputs instead.
        """

        if return_dict and not tokenize:
            raise ValueError(
                "`return_dict=True` is incompatible with `tokenize=False`, because there is no dict "
                "of tokenizer outputs to return."
            )

        if return_assistant_tokens_mask and not return_dict:
            raise ValueError("`return_assistant_tokens_mask=True` is incompatible with `return_dict=False`")

        if tokenizer_kwargs is None:
            tokenizer_kwargs = {}

        chat_template = self.get_chat_template(chat_template, tools)

        if return_assistant_tokens_mask and not re.search(r"\{\%-?\s*generation\s*-?\%\}", chat_template):
            logger.warning_once(
                "return_assistant_tokens_mask==True but chat template does not contain `{% generation %}` keyword."
            )

        # Compilation function uses a cache to avoid recompiling the same template
        compiled_template = self._compile_jinja_template(chat_template)

        if isinstance(conversation, (list, tuple)) and (
            isinstance(conversation[0], (list, tuple)) or hasattr(conversation[0], "messages")
        ):
            conversations = conversation
            is_batched = True
        else:
            conversations = [conversation]
            is_batched = False

        # We accept either JSON schemas or functions for tools. If we get functions, we convert them to schemas
        if tools is not None:
            tool_schemas = []
            for tool in tools:
                if isinstance(tool, dict):
                    tool_schemas.append(tool)
                elif isfunction(tool):
                    tool_schemas.append(get_json_schema(tool))
                else:
                    raise ValueError(
                        "Tools should either be a JSON schema, or a callable function with type hints "
                        "and a docstring suitable for auto-conversion to a schema."
                    )
        else:
            tool_schemas = None

        if documents is not None:
            for document in documents:
                if not isinstance(document, dict):
                    raise TypeError("Documents should be a list of dicts with 'title' and 'text' keys!")

        rendered = []
        all_generation_indices = []
        template_kwargs = {**self.special_tokens_map, **kwargs}  # kwargs overwrite special tokens if both are present
        for chat in conversations:
            if hasattr(chat, "messages"):
                # Indicates it's a Conversation object
                chat = chat.messages
            if return_assistant_tokens_mask:
                rendered_chat, generation_indices = self._render_with_assistant_indices(
                    compiled_template=compiled_template,
                    messages=chat,
                    tools=tool_schemas,
                    documents=documents,
                    add_generation_prompt=add_generation_prompt,
                    **template_kwargs,
                )
                all_generation_indices.append(generation_indices)
            else:
                rendered_chat = compiled_template.render(
                    messages=chat,
                    tools=tool_schemas,
                    documents=documents,
                    add_generation_prompt=add_generation_prompt,
                    **template_kwargs,
                )
            rendered.append(rendered_chat)

        if not is_batched:
            rendered = rendered[0]

        if tokenize:
            out = self(
                rendered,
                padding=padding,
                truncation=truncation,
                max_length=max_length,
                add_special_tokens=False,
                return_tensors=return_tensors,
                **tokenizer_kwargs,
            )
            if return_dict:
                if return_assistant_tokens_mask:
                    assistant_masks = []
                    if is_batched or return_tensors:
                        input_ids = out["input_ids"]
                    else:
                        input_ids = [out["input_ids"]]
                    for i in range(len(input_ids)):
                        current_mask = [0] * len(input_ids[i])
                        for assistant_start_char, assistant_end_char in all_generation_indices[i]:
                            start_token = out.char_to_token(i, assistant_start_char)
                            end_token = out.char_to_token(i, assistant_end_char - 1)
                            if start_token is None:
                                # start_token is out of bounds maybe due to truncation.
                                break
                            for token_id in range(start_token, end_token + 1 if end_token else len(input_ids)):
                                current_mask[token_id] = 1
                        assistant_masks.append(current_mask)
                    out["assistant_masks"] = assistant_masks if is_batched else assistant_masks[0]
                return out
            else:
                return out["input_ids"]
        else:
            return rendered

    def _render_with_assistant_indices(
        self, compiled_template, messages, tools, documents, add_generation_prompt, **template_kwargs
    ):
        rendered_blocks = []
        generation_indices = []
        with compiled_template.environment.activate_tracker(rendered_blocks, generation_indices):
            for block in compiled_template.generate(
                messages=messages,
                tools=tools,
                documents=documents,
                add_generation_prompt=add_generation_prompt,
                **template_kwargs,
            ):
                rendered_blocks.append(block)
            rendered_chat = "".join(rendered_blocks)
        return rendered_chat, generation_indices

    @lru_cache
    def _compile_jinja_template(self, chat_template):
        try:
            import jinja2
            from jinja2 import nodes
            from jinja2.exceptions import TemplateError
            from jinja2.ext import Extension
            from jinja2.sandbox import ImmutableSandboxedEnvironment
        except ImportError:
            raise ImportError("apply_chat_template requires jinja2 to be installed.")

        if version.parse(jinja2.__version__) < version.parse("3.1.0"):
            raise ImportError(
                "apply_chat_template requires jinja2>=3.1.0 to be installed. Your version is " f"{jinja2.__version__}."
            )

        def raise_exception(message):
            raise TemplateError(message)

        def tojson(x, ensure_ascii=False, indent=None, separators=None, sort_keys=False):
            # We override the built-in tojson filter because Jinja's default filter escapes HTML characters
            # We also expose some options like custom indents and separators
            return json.dumps(x, ensure_ascii=ensure_ascii, indent=indent, separators=separators, sort_keys=sort_keys)

        class AssistantTracker(Extension):
            # This extension is used to track the indices of assistant-generated tokens in the rendered chat
            tags = {"generation"}

            def __init__(self, environment: ImmutableSandboxedEnvironment):
                # The class is only initiated by jinja.
                super().__init__(environment)
                environment.extend(activate_tracker=self.activate_tracker)
                self._rendered_blocks = None
                self._generation_indices = None

            def parse(self, parser: jinja2.parser.Parser) -> jinja2.nodes.CallBlock:
                lineno = next(parser.stream).lineno
                body = parser.parse_statements(["name:endgeneration"], drop_needle=True)
                return nodes.CallBlock(self.call_method("_generation_support"), [], [], body).set_lineno(lineno)

            @jinja2.pass_eval_context
            def _generation_support(self, context: jinja2.nodes.EvalContext, caller: jinja2.runtime.Macro) -> str:
                rv = caller()
                if self.is_active():
                    # Only track generation indices if the tracker is active
                    start_index = len("".join(self._rendered_blocks))
                    end_index = start_index + len(rv)
                    self._generation_indices.append((start_index, end_index))
                return rv

            def is_active(self) -> bool:
                return self._rendered_blocks or self._generation_indices

            @contextmanager
            def activate_tracker(self, rendered_blocks: List[int], generation_indices: List[int]):
                try:
                    if self.is_active():
                        raise ValueError("AssistantTracker should not be reused before closed")
                    self._rendered_blocks = rendered_blocks
                    self._generation_indices = generation_indices

                    yield
                finally:
                    self._rendered_blocks = None
                    self._generation_indices = None

        jinja_env = ImmutableSandboxedEnvironment(trim_blocks=True, lstrip_blocks=True, extensions=[AssistantTracker])
        jinja_env.filters["tojson"] = tojson
        jinja_env.globals["raise_exception"] = raise_exception
        return jinja_env.from_string(chat_template)

    def get_chat_template(self, chat_template: Optional[str] = None, tools: Optional[List[Dict]] = None) -> str:
        """
        Retrieve the chat template string used for tokenizing chat messages. This template is used
        internally by the `apply_chat_template` method and can also be used externally to retrieve the model's chat
        template for better generation tracking.

        Args:
            chat_template (`str`, *optional*):
                A Jinja template or the name of a template to use for this conversion.
                It is usually not necessary to pass anything to this argument,
                as the model's template will be used by default.
            tools (`List[Dict]`, *optional*):
                A list of tools (callable functions) that will be accessible to the model. If the template does not
                support function calling, this argument will have no effect. Each tool should be passed as a JSON Schema,
                giving the name, description and argument types for the tool. See our
                [chat templating guide](https://huggingface.co/docs/transformers/main/en/chat_templating#automated-function-conversion-for-tool-use)
                for more information.

        Returns:
            `str`: The chat template string.
        """
        # First, handle the cases when the model has a dict of multiple templates
        if isinstance(self.chat_template, dict):
            template_dict = self.chat_template
            if chat_template is not None and chat_template in template_dict:
                # The user can pass the name of a template to the chat template argument instead of an entire template
                chat_template = template_dict[chat_template]
            elif chat_template is None:
                if tools is not None and "tool_use" in template_dict:
                    chat_template = template_dict["tool_use"]
                elif "default" in template_dict:
                    chat_template = template_dict["default"]
                else:
                    raise ValueError(
                        "This model has multiple chat templates with no default specified! Please either pass a chat "
                        "template or the name of the template you wish to use to the `chat_template` argument. Available "
                        f"template names are {sorted(template_dict.keys())}."
                    )

        elif chat_template is None:
            # These are the cases when the model has a single template
            # priority: `chat_template` argument > `tokenizer.chat_template`
            if self.chat_template is not None:
                chat_template = self.chat_template

            else:
                raise ValueError(
                    "Cannot use apply_chat_template() because tokenizer.chat_template is not set and no template "
                    "argument was passed! For information about writing templates and setting the "
                    "tokenizer.chat_template attribute, please see the documentation at "
                    "https://huggingface.co/docs/transformers/main/en/chat_templating"
                )

        return chat_template

    @classmethod
    def from_pretrained(
        cls,
        pretrained_model_name_or_path: Union[str, os.PathLike],
        *init_inputs,
        cache_dir: Optional[Union[str, os.PathLike]] = None,
        force_download: bool = False,
        local_files_only: bool = False,
        token: Optional[Union[str, bool]] = None,
        revision: str = "main",
        trust_remote_code=False,
        **kwargs,
    ):
        r"""
        Instantiate a [`~tokenization_utils_base.PreTrainedTokenizerBase`] (or a derived class) from a predefined
        tokenizer.

        Args:
            pretrained_model_name_or_path (`str` or `os.PathLike`):
                Can be either:

                - A string, the *model id* of a predefined tokenizer hosted inside a model repo on huggingface.co.
                - A path to a *directory* containing vocabulary files required by the tokenizer, for instance saved
                  using the [`~tokenization_utils_base.PreTrainedTokenizerBase.save_pretrained`] method, e.g.,
                  `./my_model_directory/`.
                - (**Deprecated**, not applicable to all derived classes) A path or url to a single saved vocabulary
                  file (if and only if the tokenizer only requires a single vocabulary file like Bert or XLNet), e.g.,
                  `./my_model_directory/vocab.txt`.
            cache_dir (`str` or `os.PathLike`, *optional*):
                Path to a directory in which a downloaded predefined tokenizer vocabulary files should be cached if the
                standard cache should not be used.
            force_download (`bool`, *optional*, defaults to `False`):
                Whether or not to force the (re-)download the vocabulary files and override the cached versions if they
                exist.
            resume_download:
                Deprecated and ignored. All downloads are now resumed by default when possible.
                Will be removed in v5 of Transformers.
            proxies (`Dict[str, str]`, *optional*):
                A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
                'http://hostname': 'foo.bar:4012'}`. The proxies are used on each request.
            token (`str` or *bool*, *optional*):
                The token to use as HTTP bearer authorization for remote files. If `True`, will use the token generated
                when running `huggingface-cli login` (stored in `~/.huggingface`).
            local_files_only (`bool`, *optional*, defaults to `False`):
                Whether or not to only rely on local files and not to attempt to download any files.
            revision (`str`, *optional*, defaults to `"main"`):
                The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a
                git-based system for storing models and other artifacts on huggingface.co, so `revision` can be any
                identifier allowed by git.
            subfolder (`str`, *optional*):
                In case the relevant files are located inside a subfolder of the model repo on huggingface.co (e.g. for
                facebook/rag-token-base), specify it here.
            inputs (additional positional arguments, *optional*):
                Will be passed along to the Tokenizer `__init__` method.
            trust_remote_code (`bool`, *optional*, defaults to `False`):
                Whether or not to allow for custom models defined on the Hub in their own modeling files. This option
                should only be set to `True` for repositories you trust and in which you have read the code, as it will
                execute code present on the Hub on your local machine.
            kwargs (additional keyword arguments, *optional*):
                Will be passed to the Tokenizer `__init__` method. Can be used to set special tokens like `bos_token`,
                `eos_token`, `unk_token`, `sep_token`, `pad_token`, `cls_token`, `mask_token`,
                `additional_special_tokens`. See parameters in the `__init__` for more details.

        <Tip>

        Passing `token=True` is required when you want to use a private model.

        </Tip>

        Examples:

        ```python
        # We can't instantiate directly the base class *PreTrainedTokenizerBase* so let's show our examples on a derived class: BertTokenizer
        # Download vocabulary from huggingface.co and cache.
        tokenizer = BertTokenizer.from_pretrained("google-bert/bert-base-uncased")

        # Download vocabulary from huggingface.co (user-uploaded) and cache.
        tokenizer = BertTokenizer.from_pretrained("dbmdz/bert-base-german-cased")

        # If vocabulary files are in a directory (e.g. tokenizer was saved using *save_pretrained('./test/saved_model/')*)
        tokenizer = BertTokenizer.from_pretrained("./test/saved_model/")

        # If the tokenizer uses a single vocabulary file, you can point directly to this file
        tokenizer = BertTokenizer.from_pretrained("./test/saved_model/my_vocab.txt")

        # You can link tokens to special vocabulary when instantiating
        tokenizer = BertTokenizer.from_pretrained("google-bert/bert-base-uncased", unk_token="<unk>")
        # You should be sure '<unk>' is in the vocabulary when doing that.
        # Otherwise use tokenizer.add_special_tokens({'unk_token': '<unk>'}) instead)
        assert tokenizer.unk_token == "<unk>"
        ```"""
        resume_download = kwargs.pop("resume_download", None)
        proxies = kwargs.pop("proxies", None)
        use_auth_token = kwargs.pop("use_auth_token", None)
        subfolder = kwargs.pop("subfolder", None)
        from_pipeline = kwargs.pop("_from_pipeline", None)
        from_auto_class = kwargs.pop("_from_auto", False)
        commit_hash = kwargs.pop("_commit_hash", None)
        gguf_file = kwargs.get("gguf_file", None)
        mirror = kwargs.get("mirror", "huggingface")

        if use_auth_token is not None:
            warnings.warn(
                "The `use_auth_token` argument is deprecated. Please use `token` instead.",
                FutureWarning,
            )
            if token is not None:
                raise ValueError(
                    "`token` and `use_auth_token` are both specified. Please set only the argument `token`."
                )
            token = use_auth_token

        user_agent = {"file_type": "tokenizer", "from_auto_class": from_auto_class, "is_fast": "Fast" in cls.__name__}
        if from_pipeline is not None:
            user_agent["using_pipeline"] = from_pipeline

        if is_offline_mode() and not local_files_only:
            logger.info("Offline mode: forcing local_files_only=True")
            local_files_only = True

        pretrained_model_name_or_path = str(pretrained_model_name_or_path)
        vocab_files = {}
        init_configuration = {}

        is_local = os.path.isdir(pretrained_model_name_or_path)
        single_file_id = None
        if os.path.isfile(pretrained_model_name_or_path) or is_remote_url(pretrained_model_name_or_path):
            if len(cls.vocab_files_names) > 1 and not gguf_file:
                raise ValueError(
                    f"Calling {cls.__name__}.from_pretrained() with the path to a single file or url is not "
                    "supported for this tokenizer. Use a model identifier or the path to a directory instead."
                )
            warnings.warn(
                f"Calling {cls.__name__}.from_pretrained() with the path to a single file or url is deprecated and "
                "won't be possible anymore in v5. Use a model identifier or the path to a directory instead.",
                FutureWarning,
            )
            file_id = list(cls.vocab_files_names.keys())[0]

            vocab_files[file_id] = pretrained_model_name_or_path
            single_file_id = file_id
        else:
            if gguf_file:
                vocab_files["vocab_file"] = gguf_file
            else:
                # At this point pretrained_model_name_or_path is either a directory or a model identifier name
                additional_files_names = {
                    "added_tokens_file": ADDED_TOKENS_FILE,  # kept only for legacy
                    "special_tokens_map_file": SPECIAL_TOKENS_MAP_FILE,  # kept only for legacy
                    "tokenizer_config_file": TOKENIZER_CONFIG_FILE,
                    # tokenizer_file used to initialize a slow from a fast. Properly copy the `addedTokens` instead of adding in random orders
                    "tokenizer_file": FULL_TOKENIZER_FILE,
                }
                vocab_files = {**cls.vocab_files_names, **additional_files_names}
                if "tokenizer_file" in vocab_files:
                    # Try to get the tokenizer config to see if there are versioned tokenizer files.
                    fast_tokenizer_file = FULL_TOKENIZER_FILE
                    resolved_config_file = cached_file(
                        pretrained_model_name_or_path,
                        TOKENIZER_CONFIG_FILE,
                        cache_dir=cache_dir,
                        force_download=force_download,
                        resume_download=resume_download,
                        proxies=proxies,
                        token=token,
                        revision=revision,
                        local_files_only=local_files_only,
                        subfolder=subfolder,
                        mirror=mirror,
                        user_agent=user_agent,
                        _raise_exceptions_for_gated_repo=False,
                        _raise_exceptions_for_missing_entries=False,
                        _raise_exceptions_for_connection_errors=False,
                    )
                    commit_hash = extract_commit_hash(resolved_config_file, commit_hash)
                    if resolved_config_file is not None:
                        with open(resolved_config_file, encoding="utf-8") as reader:
                            tokenizer_config = json.load(reader)
                            if "fast_tokenizer_files" in tokenizer_config:
                                fast_tokenizer_file = get_fast_tokenizer_file(tokenizer_config["fast_tokenizer_files"])
                    vocab_files["tokenizer_file"] = fast_tokenizer_file

        # Get files from url, cache, or disk depending on the case
        resolved_vocab_files = {}
        unresolved_files = []
        for file_id, file_path in vocab_files.items():
            if file_path is None:
                resolved_vocab_files[file_id] = None
            elif single_file_id == file_id:
                if os.path.isfile(file_path):
                    resolved_vocab_files[file_id] = file_path
                elif is_remote_url(file_path):
                    resolved_vocab_files[file_id] = download_url(file_path, proxies=proxies)
            else:
                resolved_vocab_files[file_id] = cached_file(
                    pretrained_model_name_or_path,
                    file_path,
                    cache_dir=cache_dir,
                    force_download=force_download,
                    proxies=proxies,
                    resume_download=resume_download,
                    local_files_only=local_files_only,
                    token=token,
                    user_agent=user_agent,
                    revision=revision,
                    subfolder=subfolder,
                    mirror=mirror,
                    _raise_exceptions_for_gated_repo=False,
                    _raise_exceptions_for_missing_entries=False,
                    _raise_exceptions_for_connection_errors=False,
                    # _commit_hash=commit_hash,
                )
                commit_hash = extract_commit_hash(resolved_vocab_files[file_id], commit_hash)

        if len(unresolved_files) > 0:
            logger.info(
                f"Can't load following files from cache: {unresolved_files} and cannot check if these "
                "files are necessary for the tokenizer to operate."
            )

        # If one passes a GGUF file path to `gguf_file` there is no need for this check as the tokenizer will be
        # loaded directly from the GGUF file.
        if all(full_file_name is None for full_file_name in resolved_vocab_files.values()) and not gguf_file:
            raise EnvironmentError(
                f"Can't load tokenizer for '{pretrained_model_name_or_path}'. If you were trying to load it from "
                "'https://huggingface.co/models', make sure you don't have a local directory with the same name. "
                f"Otherwise, make sure '{pretrained_model_name_or_path}' is the correct path to a directory "
                f"containing all relevant files for a {cls.__name__} tokenizer."
            )

        for file_id, file_path in vocab_files.items():
            if file_id not in resolved_vocab_files:
                continue

            if is_local:
                logger.info(f"loading file {file_path}")
            else:
                logger.info(f"loading file {file_path} from cache at {resolved_vocab_files[file_id]}")

        return cls._from_pretrained(
            resolved_vocab_files,
            pretrained_model_name_or_path,
            init_configuration,
            *init_inputs,
            token=token,
            cache_dir=cache_dir,
            local_files_only=local_files_only,
            _commit_hash=commit_hash,
            _is_local=is_local,
            trust_remote_code=trust_remote_code,
            **kwargs,
        )

    @classmethod
    def _from_pretrained(
        cls,
        resolved_vocab_files,
        pretrained_model_name_or_path,
        init_configuration,
        *init_inputs,
        token=None,
        cache_dir=None,
        local_files_only=False,
        _commit_hash=None,
        _is_local=False,
        trust_remote_code=False,
        **kwargs,
    ):
        # We instantiate fast tokenizers based on a slow tokenizer if we don't have access to the tokenizer.json
        # file or if `from_slow` is set to True.
        from_slow = kwargs.get("from_slow", False)
        gguf_file = kwargs.get("gguf_file", None)
        has_tokenizer_file = resolved_vocab_files.get("tokenizer_file", None) is not None

        # If one passes a GGUF file path to `gguf_file` there is no need for this check as the tokenizer will be
        # loaded directly from the GGUF file.
        if (from_slow or not has_tokenizer_file) and cls.slow_tokenizer_class is not None and not gguf_file:
            slow_tokenizer = (cls.slow_tokenizer_class)._from_pretrained(
                copy.deepcopy(resolved_vocab_files),
                pretrained_model_name_or_path,
                copy.deepcopy(init_configuration),
                *init_inputs,
                token=token,
                cache_dir=cache_dir,
                local_files_only=local_files_only,
                _commit_hash=_commit_hash,
                **(copy.deepcopy(kwargs)),
            )
        else:
            slow_tokenizer = None

        # Prepare tokenizer initialization kwargs
        # Did we saved some inputs and kwargs to reload ?
        tokenizer_config_file = resolved_vocab_files.pop("tokenizer_config_file", None)
        if tokenizer_config_file is not None:
            with open(tokenizer_config_file, encoding="utf-8") as tokenizer_config_handle:
                init_kwargs = json.load(tokenizer_config_handle)
            # First attempt. We get tokenizer_class from tokenizer_config to check mismatch between tokenizers.
            config_tokenizer_class = init_kwargs.get("tokenizer_class")
            init_kwargs.pop("tokenizer_class", None)
            if not has_tokenizer_file:
                init_kwargs.pop("tokenizer_file", None)
            saved_init_inputs = init_kwargs.pop("init_inputs", ())
            if not init_inputs:
                init_inputs = saved_init_inputs
        else:
            config_tokenizer_class = None
            init_kwargs = init_configuration


        if config_tokenizer_class is None:
            # Matt: This entire block is only used to decide if the tokenizer class matches the class in the repo.
            #       If not, it raises a warning, but otherwise continues. Since we mostly load tokenizers with
            #       AutoTokenizer these days, it seems like a lot of work (and a source of bugs) for little gain.
            #       Maybe we can just remove this entirely?
            from .models.auto.configuration_auto import AutoConfig  # tests_ignore

            # Second attempt. If we have not yet found tokenizer_class, let's try to use the config.
            try:
                config = AutoConfig.from_pretrained(
                    pretrained_model_name_or_path,
                    token=token,
                    cache_dir=cache_dir,
                    local_files_only=local_files_only,
                    trust_remote_code=trust_remote_code,
                    _commit_hash=_commit_hash,
                )
                config_tokenizer_class = config.tokenizer_class
            except (OSError, ValueError, KeyError):
                # skip if an error occurred.
                config = None
            if config_tokenizer_class is None:
                # Third attempt. If we have not yet found the original type of the tokenizer,
                # we are loading we see if we can infer it from the type of the configuration file
                from .models.auto.tokenization_auto import TOKENIZER_MAPPING_NAMES  # tests_ignore

                if hasattr(config, "model_type"):
                    model_type = config.model_type
                else:
                    # Fallback: use pattern matching on the string.
                    model_type = None
                    for pattern in TOKENIZER_MAPPING_NAMES.keys():
                        if pattern in str(pretrained_model_name_or_path):
                            model_type = pattern
                            break

                if model_type is not None:
                    config_tokenizer_class, config_tokenizer_class_fast = TOKENIZER_MAPPING_NAMES.get(
                        model_type, (None, None)
                    )
                    if config_tokenizer_class is None:
                        config_tokenizer_class = config_tokenizer_class_fast

        if config_tokenizer_class is not None:
            if cls.__name__.replace("Fast", "") != config_tokenizer_class.replace("Fast", ""):
                logger.warning(
                    "The tokenizer class you load from this checkpoint is not the same type as the class this"
                    " function is called from. It may result in unexpected tokenization. \nThe tokenizer class you"
                    f" load from this checkpoint is '{config_tokenizer_class}'. \nThe class this function is called"
                    f" from is '{cls.__name__}'."
                )

        # Update with newly provided kwargs
        init_kwargs.update(kwargs)

        # Merge resolved_vocab_files arguments in init_kwargs.
        added_tokens_file = resolved_vocab_files.pop("added_tokens_file", None)
        special_tokens_map_file = resolved_vocab_files.pop("special_tokens_map_file", None)
        for args_name, file_path in resolved_vocab_files.items():
            if args_name not in init_kwargs:
                init_kwargs[args_name] = file_path
        tokenizer_file = resolved_vocab_files.pop("tokenizer_file", None)

        if slow_tokenizer is not None:
            init_kwargs["__slow_tokenizer"] = slow_tokenizer
        init_kwargs["name_or_path"] = pretrained_model_name_or_path

        #### Handle tokenizer serialization of added and special tokens
        added_tokens_decoder: Dict[int, AddedToken] = {}
        added_tokens_map: Dict[str, AddedToken] = {}
        # if we have info on the slow added tokens
        if "added_tokens_decoder" in init_kwargs:
            for idx, token in init_kwargs["added_tokens_decoder"].items():
                if isinstance(token, dict):
                    token = AddedToken(**token)
                if isinstance(token, AddedToken):
                    added_tokens_decoder[int(idx)] = token
                    added_tokens_map[str(token)] = token
                else:
                    raise ValueError(
                        f"Found a {token.__class__} in the saved `added_tokens_decoder`, should be a dictionary or an AddedToken instance"
                    )
        else:
            # begin legacy: read the added_tokens_file and update kwargs with special_tokens_map if modified
            if special_tokens_map_file is not None:
                with open(special_tokens_map_file, encoding="utf-8") as special_tokens_map_handle:
                    special_tokens_map = json.load(special_tokens_map_handle)
                    for key, value in special_tokens_map.items():
                        if key in kwargs and kwargs[key]:
                            # This value has already been redefined by the kwargs
                            # We keep this new value and ignore the one stored in the special_tokens_map_file
                            continue
                        if isinstance(value, dict):
                            value["special"] = True
                            value = AddedToken(**value)
                        elif key == "additional_special_tokens" and isinstance(value, list):
                            additional_special_tokens = init_kwargs.pop("additional_special_tokens", []) or []
                            for token in value:
                                if isinstance(token, dict):
                                    token["special"] = True
                                    token = AddedToken(**token)
                                if token not in additional_special_tokens:
                                    additional_special_tokens.append(token)
                            value = additional_special_tokens
                        init_kwargs[key] = value

            # slow -> slow|fast, legacy: convert the `"added_tokens.json"` file to `added_tokens_decoder`.
            # this is for legacy purpose. We don't add the tokens after init for efficiency.
            if added_tokens_file is not None:
                special_tokens = []
                for key in cls.SPECIAL_TOKENS_ATTRIBUTES & init_kwargs.keys():
                    if init_kwargs[key] is not None:
                        if key == "additional_special_tokens":
                            special_tokens += [str(token) for token in init_kwargs[key]]
                        else:
                            special_tokens.append(str(init_kwargs[key]))

                with open(added_tokens_file, encoding="utf-8") as added_tokens_handle:
                    added_tok_encoder = json.load(added_tokens_handle)
                for str_token, index in added_tok_encoder.items():
                    # if index not in added_tokens_decoder and str_token not in added_tokens_map:
                    special = str_token in special_tokens
                    added_tokens_decoder[index] = AddedToken(
                        str_token, rstrip=False, lstrip=False, normalized=not special, special=special
                    )
                    added_tokens_map[str(token)] = added_tokens_decoder[index]

            # allows converting a fast -> slow: add the `tokenizer.json`'s `"added_tokens"` to the slow tokenizer
            # if `tokenizer_config.json` is `None`
            if tokenizer_file is not None:
                # This is for slow so can be done before
                with open(tokenizer_file, encoding="utf-8") as tokenizer_file_handle:
                    tokenizer_file_handle = json.load(tokenizer_file_handle)
                    added_tokens = tokenizer_file_handle.pop("added_tokens")
                for serialized_tokens in added_tokens:
                    idx = serialized_tokens.pop("id")
                    added_tokens_decoder[idx] = AddedToken(**serialized_tokens)
                    added_tokens_map[str(added_tokens_decoder[idx])] = added_tokens_decoder[idx]
            # end legacy

        # Passing AddedTokens and not strings to the class to prevent it from casting the string to a different AddedToken
        # convert {'__type': 'AddedToken', 'content': '<ent>', 'lstrip': False, 'normalized': True, ...} to AddedTokens
        init_kwargs["added_tokens_decoder"] = added_tokens_decoder
        init_kwargs = cls.convert_added_tokens(init_kwargs, save=False)
        for key in cls.SPECIAL_TOKENS_ATTRIBUTES & init_kwargs.keys():
            if added_tokens_map and init_kwargs[key] is not None:
                if key != "additional_special_tokens":
                    init_kwargs[key] = added_tokens_map.get(str(init_kwargs[key]), init_kwargs[key])

        # Instantiate the tokenizer.
        try:
            tokenizer = cls(*init_inputs, **init_kwargs)
        except OSError:
            raise OSError(
                "Unable to load vocabulary from file. "
                "Please check that the provided vocabulary is accessible and not corrupted."
            )

        if added_tokens_decoder and max(list(added_tokens_decoder.keys())[-1], 0) > tokenizer.vocab_size:
            logger.info(
                "Special tokens have been added in the vocabulary, make sure the associated word embeddings are"
                " fine-tuned or trained."
            )
        return tokenizer

    @staticmethod
    def _eventually_correct_t5_max_length(pretrained_model_name_or_path, max_model_length, init_max_model_length):
        # This method should be deleted in Transformers v5
        # Its only purpose is to potentially throw a warning
        # that incorrectly defined max lengths of T5's tokenizer are used
        # which we will correct in Transformers v5.
        return max_model_length

    @classmethod
    def convert_added_tokens(cls, obj: Union[AddedToken, Any], save=False, add_type_field=True):
        if isinstance(obj, dict) and "__type" in obj and obj["__type"] == "AddedToken":
            obj.pop("__type")
            return AddedToken(**obj)
        if isinstance(obj, AddedToken) and save:
            obj = obj.__getstate__()
            if add_type_field:
                obj["__type"] = "AddedToken"
            else:
                # Don't save "special" for previous tokenizers
                obj.pop("special")
            return obj
        elif isinstance(obj, (list, tuple)):
            return [cls.convert_added_tokens(o, save=save, add_type_field=add_type_field) for o in obj]
        elif isinstance(obj, dict):
            return {k: cls.convert_added_tokens(v, save=save, add_type_field=add_type_field) for k, v in obj.items()}
        return obj

    def save_pretrained(
        self,
        save_directory: Union[str, os.PathLike],
        legacy_format: Optional[bool] = None,
        filename_prefix: Optional[str] = None,
        push_to_hub: bool = False,
        **kwargs,
    ) -> Tuple[str]:
        """
        Save the full tokenizer state.


        This method make sure the full tokenizer can then be re-loaded using the
        [`~tokenization_utils_base.PreTrainedTokenizer.from_pretrained`] class method..

        Warning,None This won't save modifications you may have applied to the tokenizer after the instantiation (for
        instance, modifying `tokenizer.do_lower_case` after creation).

        Args:
            save_directory (`str` or `os.PathLike`): The path to a directory where the tokenizer will be saved.
            legacy_format (`bool`, *optional*):
                Only applicable for a fast tokenizer. If unset (default), will save the tokenizer in the unified JSON
                format as well as in legacy format if it exists, i.e. with tokenizer specific vocabulary and a separate
                added_tokens files.

                If `False`, will only save the tokenizer in the unified JSON format. This format is incompatible with
                "slow" tokenizers (not powered by the *tokenizers* library), so the tokenizer will not be able to be
                loaded in the corresponding "slow" tokenizer.

                If `True`, will save the tokenizer in legacy format. If the "slow" tokenizer doesn't exits, a value
                error is raised.
            filename_prefix (`str`, *optional*):
                A prefix to add to the names of the files saved by the tokenizer.
            push_to_hub (`bool`, *optional*, defaults to `False`):
                Whether or not to push your model to the Hugging Face model hub after saving it. You can specify the
                repository you want to push to with `repo_id` (will default to the name of `save_directory` in your
                namespace).
            kwargs (`Dict[str, Any]`, *optional*):
                Additional key word arguments passed along to the [`~utils.PushToHubMixin.push_to_hub`] method.

        Returns:
            A tuple of `str`: The files saved.
        """
        use_auth_token = kwargs.pop("use_auth_token", None)

        if use_auth_token is not None:
            warnings.warn(
                "The `use_auth_token` argument is deprecated. Please use `token` instead.",
                FutureWarning,
            )
            if kwargs.get("token", None) is not None:
                raise ValueError(
                    "`token` and `use_auth_token` are both specified. Please set only the argument `token`."
                )
            kwargs["token"] = use_auth_token

        if os.path.isfile(save_directory):
            logger.error(f"Provided path ({save_directory}) should be a directory, not a file")
            return

        os.makedirs(save_directory, exist_ok=True)

        if push_to_hub:
            commit_message = kwargs.pop("commit_message", None)
            repo_id = kwargs.pop("repo_id", save_directory.split(os.path.sep)[-1])
            repo_id = self._create_repo(repo_id, **kwargs)
            files_timestamps = self._get_files_timestamps(save_directory)

        special_tokens_map_file = os.path.join(
            save_directory, (filename_prefix + "-" if filename_prefix else "") + SPECIAL_TOKENS_MAP_FILE
        )
        tokenizer_config_file = os.path.join(
            save_directory, (filename_prefix + "-" if filename_prefix else "") + TOKENIZER_CONFIG_FILE
        )

        tokenizer_config = copy.deepcopy(self.init_kwargs)

        # Let's save the init kwargs
        target_keys = set(self.init_kwargs.keys())
        # Let's save the special tokens map (only the strings)
        target_keys.update(["model_max_length", "clean_up_tokenization_spaces"])

        for k in target_keys:
            if hasattr(self, k):
                tokenizer_config[k] = getattr(self, k)

        # Let's make sure we properly save the special tokens.
        tokenizer_config.update(self.special_tokens_map)

        if self.chat_template is not None:
            if isinstance(self.chat_template, dict):
                # Chat template dicts are saved to the config as lists of dicts with fixed key names.
                # They will be reconstructed as a single dict during loading.
                tokenizer_config["chat_template"] = [{"name": k, "template": v} for k, v in self.chat_template.items()]
            else:
                tokenizer_config["chat_template"] = self.chat_template

        if len(self.init_inputs) > 0:
            tokenizer_config["init_inputs"] = copy.deepcopy(self.init_inputs)
        for file_id in self.vocab_files_names.keys():
            tokenizer_config.pop(file_id, None)

        # no typefields, this way old fast and slow can load it
        tokenizer_config = self.convert_added_tokens(tokenizer_config, add_type_field=True, save=True)

        # Process added tokens seperatly: allows previous versions to ignore it!
        added_tokens = {}
        for key, value in self.added_tokens_decoder.items():
            added_tokens[key] = value.__getstate__()
        tokenizer_config["added_tokens_decoder"] = added_tokens

        # Add tokenizer class to the tokenizer config to be able to reload it with from_pretrained
        tokenizer_class = self.__class__.__name__
        # Remove the Fast at the end unless we have a special `PreTrainedTokenizerFast`
        if tokenizer_class.endswith("Fast") and tokenizer_class != "PreTrainedTokenizerFast":
            tokenizer_class = tokenizer_class[:-4]
        tokenizer_config["tokenizer_class"] = tokenizer_class
        if getattr(self, "_auto_map", None) is not None:
            tokenizer_config["auto_map"] = self._auto_map
        if getattr(self, "_processor_class", None) is not None:
            tokenizer_config["processor_class"] = self._processor_class

        # remove private information
        if "name_or_path" in tokenizer_config:
            tokenizer_config.pop("name_or_path")
            tokenizer_config.pop("special_tokens_map_file", None)
            tokenizer_config.pop("tokenizer_file", None)
        if "device_map" in tokenizer_config:
            tokenizer_config.pop("device_map")

        with open(tokenizer_config_file, "w", encoding="utf-8") as f:
            out_str = json.dumps(tokenizer_config, indent=2, sort_keys=True, ensure_ascii=False) + "\n"
            f.write(out_str)
        logger.info(f"tokenizer config file saved in {tokenizer_config_file}")

        # Sanitize AddedTokens in special_tokens_map

        # kept for forward compatibility, will be removed in transoformers 5. Typefields are not saved for FC, special should not be save either
        write_dict = self.convert_added_tokens(self.special_tokens_map_extended, save=True, add_type_field=False)
        with open(special_tokens_map_file, "w", encoding="utf-8") as f:
            out_str = json.dumps(write_dict, indent=2, sort_keys=True, ensure_ascii=False) + "\n"
            f.write(out_str)
        logger.info(f"Special tokens file saved in {special_tokens_map_file}")

        file_names = (tokenizer_config_file, special_tokens_map_file)

        save_files = self._save_pretrained(
            save_directory=save_directory,
            file_names=file_names,
            legacy_format=legacy_format,
            filename_prefix=filename_prefix,
        )

        if push_to_hub:
            self._upload_modified_files(
                save_directory,
                repo_id,
                files_timestamps,
                commit_message=commit_message,
                token=kwargs.get("token"),
            )

        return save_files

    def _save_pretrained(
        self,
        save_directory: Union[str, os.PathLike],
        file_names: Tuple[str],
        legacy_format: Optional[bool] = None,
        filename_prefix: Optional[str] = None,
    ) -> Tuple[str]:
        """
        Save a tokenizer using the slow-tokenizer/legacy format: vocabulary + added tokens.

        Fast tokenizers can also be saved in a unique JSON file containing {config + vocab + added-tokens} using the
        specific [`~tokenization_utils_fast.PreTrainedTokenizerFast._save_pretrained`]
        """
        if legacy_format is False:
            raise ValueError(
                "Only fast tokenizers (instances of PreTrainedTokenizerFast) can be saved in non legacy format."
            )

        save_directory = str(save_directory)

        added_tokens_file = os.path.join(
            save_directory, (filename_prefix + "-" if filename_prefix else "") + ADDED_TOKENS_FILE
        )
        # the new get_added_vocab() also returns special tokens and tokens that have an index < vocab_size
        added_vocab = {tok: index for tok, index in self.added_tokens_encoder.items() if index >= self.vocab_size}
        if added_vocab:
            with open(added_tokens_file, "w", encoding="utf-8") as f:
                out_str = json.dumps(added_vocab, indent=2, sort_keys=True, ensure_ascii=False) + "\n"
                f.write(out_str)
                logger.info(f"added tokens file saved in {added_tokens_file}")

        vocab_files = self.save_vocabulary(save_directory, filename_prefix=filename_prefix)

        return file_names + vocab_files + (added_tokens_file,)

    def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
        """
        Save only the vocabulary of the tokenizer (vocabulary + added tokens).

        This method won't save the configuration and special token mappings of the tokenizer. Use
        [`~PreTrainedTokenizerFast._save_pretrained`] to save the whole state of the tokenizer.

        Args:
            save_directory (`str`):
                The directory in which to save the vocabulary.
            filename_prefix (`str`, *optional*):
                An optional prefix to add to the named of the saved files.

        Returns:
            `Tuple(str)`: Paths to the files saved.
        """
        raise NotImplementedError

    def tokenize(self, text: str, pair: Optional[str] = None, add_special_tokens: bool = False, **kwargs) -> List[str]:
        """
        Converts a string into a sequence of tokens, replacing unknown tokens with the `unk_token`.

        Args:
            text (`str`):
                The sequence to be encoded.
            pair (`str`, *optional*):
                A second sequence to be encoded with the first.
            add_special_tokens (`bool`, *optional*, defaults to `False`):
                Whether or not to add the special tokens associated with the corresponding model.
            kwargs (additional keyword arguments, *optional*):
                Will be passed to the underlying model specific encode method. See details in
                [`~PreTrainedTokenizerBase.__call__`]

        Returns:
            `List[str]`: The list of tokens.
        """
        raise NotImplementedError

    def encode(
        self,
        text: Union[TextInput, PreTokenizedInput, EncodedInput],
        text_pair: Optional[Union[TextInput, PreTokenizedInput, EncodedInput]] = None,
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        stride: int = 0,
        return_tensors: Optional[Union[str, TensorType]] = None,
        **kwargs,
    ) -> List[int]:
        """
        Converts a string to a sequence of ids (integer), using the tokenizer and vocabulary.

        Same as doing `self.convert_tokens_to_ids(self.tokenize(text))`.

        Args:
            text (`str`, `List[str]` or `List[int]`):
                The first sequence to be encoded. This can be a string, a list of strings (tokenized string using the
                `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
                method).
            text_pair (`str`, `List[str]` or `List[int]`, *optional*):
                Optional second sequence to be encoded. This can be a string, a list of strings (tokenized string using
                the `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
                method).
        """
        encoded_inputs = self.encode_plus(
            text,
            text_pair=text_pair,
            add_special_tokens=add_special_tokens,
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            stride=stride,
            return_tensors=return_tensors,
            **kwargs,
        )

        return encoded_inputs["input_ids"]

    def num_special_tokens_to_add(self, pair: bool = False) -> int:
        raise NotImplementedError

    def _get_padding_truncation_strategies(
        self, padding=False, truncation=None, max_length=None, pad_to_multiple_of=None, verbose=True, **kwargs
    ):
        """
        Find the correct padding/truncation strategy with backward compatibility for old arguments (truncation_strategy
        and pad_to_max_length) and behaviors.
        """
        old_truncation_strategy = kwargs.pop("truncation_strategy", "do_not_truncate")
        old_pad_to_max_length = kwargs.pop("pad_to_max_length", False)

        # Backward compatibility for previous behavior, maybe we should deprecate it:
        # If you only set max_length, it activates truncation for max_length
        if max_length is not None and padding is False and truncation is None:
            if verbose:
                if not self.deprecation_warnings.get("Truncation-not-explicitly-activated", False):
                    logger.warning(
                        "Truncation was not explicitly activated but `max_length` is provided a specific value, please"
                        " use `truncation=True` to explicitly truncate examples to max length. Defaulting to"
                        " 'longest_first' truncation strategy. If you encode pairs of sequences (GLUE-style) with the"
                        " tokenizer you can select this strategy more precisely by providing a specific strategy to"
                        " `truncation`."
                    )
                self.deprecation_warnings["Truncation-not-explicitly-activated"] = True
            truncation = "longest_first"

        # Get padding strategy
        if padding is False and old_pad_to_max_length:
            if verbose:
                warnings.warn(
                    "The `pad_to_max_length` argument is deprecated and will be removed in a future version, "
                    "use `padding=True` or `padding='longest'` to pad to the longest sequence in the batch, or "
                    "use `padding='max_length'` to pad to a max length. In this case, you can give a specific "
                    "length with `max_length` (e.g. `max_length=45`) or leave max_length to None to pad to the "
                    "maximal input size of the model (e.g. 512 for Bert).",
                    FutureWarning,
                )
            if max_length is None:
                padding_strategy = PaddingStrategy.LONGEST
            else:
                padding_strategy = PaddingStrategy.MAX_LENGTH
        elif padding is not False:
            if padding is True:
                if verbose:
                    if max_length is not None and (
                        truncation is None or truncation is False or truncation == "do_not_truncate"
                    ):
                        warnings.warn(
                            "`max_length` is ignored when `padding`=`True` and there is no truncation strategy. "
                            "To pad to max length, use `padding='max_length'`."
                        )
                    if old_pad_to_max_length is not False:
                        warnings.warn("Though `pad_to_max_length` = `True`, it is ignored because `padding`=`True`.")
                padding_strategy = PaddingStrategy.LONGEST  # Default to pad to the longest sequence in the batch
            elif not isinstance(padding, PaddingStrategy):
                padding_strategy = PaddingStrategy(padding)
            elif isinstance(padding, PaddingStrategy):
                padding_strategy = padding
        else:
            padding_strategy = PaddingStrategy.DO_NOT_PAD

        # Get truncation strategy
        if truncation is None and old_truncation_strategy != "do_not_truncate":
            if verbose:
                warnings.warn(
                    "The `truncation_strategy` argument is deprecated and will be removed in a future version, use"
                    " `truncation=True` to truncate examples to a max length. You can give a specific length with"
                    " `max_length` (e.g. `max_length=45`) or leave max_length to None to truncate to the maximal input"
                    " size of the model (e.g. 512 for Bert).  If you have pairs of inputs, you can give a specific"
                    " truncation strategy selected among `truncation='only_first'` (will only truncate the first"
                    " sentence in the pairs) `truncation='only_second'` (will only truncate the second sentence in the"
                    " pairs) or `truncation='longest_first'` (will iteratively remove tokens from the longest sentence"
                    " in the pairs).",
                    FutureWarning,
                )
            truncation_strategy = TruncationStrategy(old_truncation_strategy)
        elif truncation is not False and truncation is not None:
            if truncation is True:
                truncation_strategy = (
                    TruncationStrategy.LONGEST_FIRST
                )  # Default to truncate the longest sequences in pairs of inputs
            elif not isinstance(truncation, TruncationStrategy):
                truncation_strategy = TruncationStrategy(truncation)
            elif isinstance(truncation, TruncationStrategy):
                truncation_strategy = truncation
        else:
            truncation_strategy = TruncationStrategy.DO_NOT_TRUNCATE

        # Set max length if needed
        if max_length is None:
            if padding_strategy == PaddingStrategy.MAX_LENGTH:
                if self.model_max_length > LARGE_INTEGER:
                    if verbose:
                        if not self.deprecation_warnings.get("Asking-to-pad-to-max_length", False):
                            logger.warning(
                                "Asking to pad to max_length but no maximum length is provided and the model has no"
                                " predefined maximum length. Default to no padding."
                            )
                        self.deprecation_warnings["Asking-to-pad-to-max_length"] = True
                    padding_strategy = PaddingStrategy.DO_NOT_PAD
                else:
                    max_length = self.model_max_length

            if truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE:
                if self.model_max_length > LARGE_INTEGER:
                    if verbose:
                        if not self.deprecation_warnings.get("Asking-to-truncate-to-max_length", False):
                            logger.warning(
                                "Asking to truncate to max_length but no maximum length is provided and the model has"
                                " no predefined maximum length. Default to no truncation."
                            )
                        self.deprecation_warnings["Asking-to-truncate-to-max_length"] = True
                    truncation_strategy = TruncationStrategy.DO_NOT_TRUNCATE
                else:
                    max_length = self.model_max_length

        # Test if we have a padding token
        if padding_strategy != PaddingStrategy.DO_NOT_PAD and (self.pad_token is None or self.pad_token_id < 0):
            raise ValueError(
                "Asking to pad but the tokenizer does not have a padding token. "
                "Please select a token to use as `pad_token` `(tokenizer.pad_token = tokenizer.eos_token e.g.)` "
                "or add a new pad token via `tokenizer.add_special_tokens({'pad_token': '[PAD]'})`."
            )

        # Check that we will truncate to a multiple of pad_to_multiple_of if both are provided
        if (
            truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE
            and padding_strategy != PaddingStrategy.DO_NOT_PAD
            and pad_to_multiple_of is not None
            and max_length is not None
            and (max_length % pad_to_multiple_of != 0)
        ):
            raise ValueError(
                "Truncation and padding are both activated but "
                f"truncation length ({max_length}) is not a multiple of pad_to_multiple_of ({pad_to_multiple_of})."
            )

        return padding_strategy, truncation_strategy, max_length, kwargs

    def __call__(
        self,
        text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None,
        text_pair: Optional[Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]]] = None,
        text_target: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None,
        text_pair_target: Optional[
            Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]]
        ] = None,
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: bool = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        **kwargs,
    ) -> BatchEncoding:
        """
        Main method to tokenize and prepare for the model one or several sequence(s) or one or several pair(s) of
        sequences.

        Args:
            text (`str`, `List[str]`, `List[List[str]]`, *optional*):
                The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings
                (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set
                `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
            text_pair (`str`, `List[str]`, `List[List[str]]`, *optional*):
                The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings
                (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set
                `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
            text_target (`str`, `List[str]`, `List[List[str]]`, *optional*):
                The sequence or batch of sequences to be encoded as target texts. Each sequence can be a string or a
                list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized),
                you must set `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
            text_pair_target (`str`, `List[str]`, `List[List[str]]`, *optional*):
                The sequence or batch of sequences to be encoded as target texts. Each sequence can be a string or a
                list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized),
                you must set `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
        """
        # To avoid duplicating
        all_kwargs = {
            "add_special_tokens": add_special_tokens,
            "padding": padding,
            "truncation": truncation,
            "max_length": max_length,
            "stride": stride,
            "is_split_into_words": is_split_into_words,
            "pad_to_multiple_of": pad_to_multiple_of,
            "return_tensors": return_tensors,
            "return_token_type_ids": return_token_type_ids,
            "return_attention_mask": return_attention_mask,
            "return_overflowing_tokens": return_overflowing_tokens,
            "return_special_tokens_mask": return_special_tokens_mask,
            "return_offsets_mapping": return_offsets_mapping,
            "return_length": return_length,
            "split_special_tokens": kwargs.pop("split_special_tokens", self.split_special_tokens),
            "verbose": verbose,
        }
        all_kwargs.update(kwargs)
        if text is None and text_target is None:
            raise ValueError("You need to specify either `text` or `text_target`.")
        if text is not None:
            # for mindspore.dataset
            if isinstance(text, np.ndarray):
                text = str(text)
                if isinstance(text_pair, np.ndarray):
                    text_pair = str(text_pair)
                elif isinstance(text_pair, list):
                    text_pair = [str(t) for t in text_pair]
            # The context manager will send the inputs as normal texts and not text_target, but we shouldn't change the
            # input mode in this case.
            if not self._in_target_context_manager:
                self._switch_to_input_mode()
            encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs)
        if text_target is not None:
            self._switch_to_target_mode()
            target_encodings = self._call_one(text=text_target, text_pair=text_pair_target, **all_kwargs)
        # Leave back tokenizer in input mode
        self._switch_to_input_mode()

        if text_target is None:
            return encodings
        elif text is None:
            return target_encodings
        else:
            encodings["labels"] = target_encodings["input_ids"]
            return encodings

    def _call_one(
        self,
        text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]],
        text_pair: Optional[Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]]] = None,
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: bool = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        split_special_tokens: bool = False,
        **kwargs,
    ) -> BatchEncoding:
        # Input type checking for clearer error
        def _is_valid_text_input(t):
            if isinstance(t, str):
                # Strings are fine
                return True
            elif isinstance(t, (list, tuple)):
                # List are fine as long as they are...
                if len(t) == 0:
                    # ... empty
                    return True
                elif isinstance(t[0], str):
                    # ... list of strings
                    return True
                elif isinstance(t[0], (list, tuple)):
                    # ... list with an empty list or with a list of strings
                    return len(t[0]) == 0 or isinstance(t[0][0], str)
                else:
                    return False
            else:
                return False

        if not _is_valid_text_input(text):
            raise ValueError(
                "text input must be of type `str` (single example), `List[str]` (batch or single pretokenized example) "
                "or `List[List[str]]` (batch of pretokenized examples)."
            )

        if text_pair is not None and not _is_valid_text_input(text_pair):
            raise ValueError(
                "text input must be of type `str` (single example), `List[str]` (batch or single pretokenized example) "
                "or `List[List[str]]` (batch of pretokenized examples)."
            )

        if is_split_into_words:
            is_batched = isinstance(text, (list, tuple)) and text and isinstance(text[0], (list, tuple))
        else:
            is_batched = isinstance(text, (list, tuple))

        if is_batched:
            if isinstance(text_pair, str):
                raise TypeError(
                    "when tokenizing batches of text, `text_pair` must be a list or tuple with the same length as"
                    " `text`."
                )
            if text_pair is not None and len(text) != len(text_pair):
                raise ValueError(
                    f"batch length of `text`: {len(text)} does not match batch length of `text_pair`:"
                    f" {len(text_pair)}."
                )
            batch_text_or_text_pairs = list(zip(text, text_pair)) if text_pair is not None else text
            return self.batch_encode_plus(
                batch_text_or_text_pairs=batch_text_or_text_pairs,
                add_special_tokens=add_special_tokens,
                padding=padding,
                truncation=truncation,
                max_length=max_length,
                stride=stride,
                is_split_into_words=is_split_into_words,
                pad_to_multiple_of=pad_to_multiple_of,
                return_tensors=return_tensors,
                return_token_type_ids=return_token_type_ids,
                return_attention_mask=return_attention_mask,
                return_overflowing_tokens=return_overflowing_tokens,
                return_special_tokens_mask=return_special_tokens_mask,
                return_offsets_mapping=return_offsets_mapping,
                return_length=return_length,
                verbose=verbose,
                split_special_tokens=split_special_tokens,
                **kwargs,
            )
        else:
            return self.encode_plus(
                text=text,
                text_pair=text_pair,
                add_special_tokens=add_special_tokens,
                padding=padding,
                truncation=truncation,
                max_length=max_length,
                stride=stride,
                is_split_into_words=is_split_into_words,
                pad_to_multiple_of=pad_to_multiple_of,
                return_tensors=return_tensors,
                return_token_type_ids=return_token_type_ids,
                return_attention_mask=return_attention_mask,
                return_overflowing_tokens=return_overflowing_tokens,
                return_special_tokens_mask=return_special_tokens_mask,
                return_offsets_mapping=return_offsets_mapping,
                return_length=return_length,
                verbose=verbose,
                split_special_tokens=split_special_tokens,
                **kwargs,
            )

    def encode_plus(
        self,
        text: Union[TextInput, PreTokenizedInput, EncodedInput],
        text_pair: Optional[Union[TextInput, PreTokenizedInput, EncodedInput]] = None,
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: bool = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        **kwargs,
    ) -> BatchEncoding:
        """
        Tokenize and prepare for the model a sequence or a pair of sequences.

        <Tip warning={true}>

        This method is deprecated, `__call__` should be used instead.

        </Tip>

        Args:
            text (`str`, `List[str]` or (for non-fast tokenizers) `List[int]`):
                The first sequence to be encoded. This can be a string, a list of strings (tokenized string using the
                `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
                method).
            text_pair (`str`, `List[str]` or `List[int]`, *optional*):
                Optional second sequence to be encoded. This can be a string, a list of strings (tokenized string using
                the `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
                method).
        """

        # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
        padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            pad_to_multiple_of=pad_to_multiple_of,
            verbose=verbose,
            **kwargs,
        )

        return self._encode_plus(
            text=text,
            text_pair=text_pair,
            add_special_tokens=add_special_tokens,
            padding_strategy=padding_strategy,
            truncation_strategy=truncation_strategy,
            max_length=max_length,
            stride=stride,
            is_split_into_words=is_split_into_words,
            pad_to_multiple_of=pad_to_multiple_of,
            return_tensors=return_tensors,
            return_token_type_ids=return_token_type_ids,
            return_attention_mask=return_attention_mask,
            return_overflowing_tokens=return_overflowing_tokens,
            return_special_tokens_mask=return_special_tokens_mask,
            return_offsets_mapping=return_offsets_mapping,
            return_length=return_length,
            verbose=verbose,
            split_special_tokens=kwargs.pop("split_special_tokens", self.split_special_tokens),
            **kwargs,
        )

    def _encode_plus(
        self,
        text: Union[TextInput, PreTokenizedInput, EncodedInput],
        text_pair: Optional[Union[TextInput, PreTokenizedInput, EncodedInput]] = None,
        add_special_tokens: bool = True,
        padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
        truncation_strategy: TruncationStrategy = TruncationStrategy.DO_NOT_TRUNCATE,
        max_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: bool = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        split_special_tokens: bool = False,
        **kwargs,
    ) -> BatchEncoding:
        raise NotImplementedError

    def batch_encode_plus(
        self,
        batch_text_or_text_pairs: Union[
            List[TextInput],
            List[TextInputPair],
            List[PreTokenizedInput],
            List[PreTokenizedInputPair],
            List[EncodedInput],
            List[EncodedInputPair],
        ],
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: bool = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        split_special_tokens: bool = False,
        **kwargs,
    ) -> BatchEncoding:
        """
        Tokenize and prepare for the model a list of sequences or a list of pairs of sequences.

        <Tip warning={true}>

        This method is deprecated, `__call__` should be used instead.

        </Tip>

        Args:
            batch_text_or_text_pairs (`List[str]`, `List[Tuple[str, str]]`, `List[List[str]]`, `List[Tuple[List[str], List[str]]]`, and for not-fast tokenizers,
            also `List[List[int]]`, `List[Tuple[List[int], List[int]]]`):
                Batch of sequences or pair of sequences to be encoded. This can be a list of
                string/string-sequences/int-sequences or a list of pair of string/string-sequences/int-sequence (see
                details in `encode_plus`).
        """

        # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
        padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            pad_to_multiple_of=pad_to_multiple_of,
            verbose=verbose,
            **kwargs,
        )

        return self._batch_encode_plus(
            batch_text_or_text_pairs=batch_text_or_text_pairs,
            add_special_tokens=add_special_tokens,
            padding_strategy=padding_strategy,
            truncation_strategy=truncation_strategy,
            max_length=max_length,
            stride=stride,
            is_split_into_words=is_split_into_words,
            pad_to_multiple_of=pad_to_multiple_of,
            return_tensors=return_tensors,
            return_token_type_ids=return_token_type_ids,
            return_attention_mask=return_attention_mask,
            return_overflowing_tokens=return_overflowing_tokens,
            return_special_tokens_mask=return_special_tokens_mask,
            return_offsets_mapping=return_offsets_mapping,
            return_length=return_length,
            verbose=verbose,
            split_special_tokens=split_special_tokens,
            **kwargs,
        )

    def _batch_encode_plus(
        self,
        batch_text_or_text_pairs: Union[
            List[TextInput],
            List[TextInputPair],
            List[PreTokenizedInput],
            List[PreTokenizedInputPair],
            List[EncodedInput],
            List[EncodedInputPair],
        ],
        add_special_tokens: bool = True,
        padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
        truncation_strategy: TruncationStrategy = TruncationStrategy.DO_NOT_TRUNCATE,
        max_length: Optional[int] = None,
        stride: int = 0,
        is_split_into_words: bool = False,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        split_special_tokens: bool = False,
        **kwargs,
    ) -> BatchEncoding:
        raise NotImplementedError

    def pad(
        self,
        encoded_inputs: Union[
            BatchEncoding,
            List[BatchEncoding],
            Dict[str, EncodedInput],
            Dict[str, List[EncodedInput]],
            List[Dict[str, EncodedInput]],
        ],
        padding: Union[bool, str, PaddingStrategy] = True,
        max_length: Optional[int] = None,
        pad_to_multiple_of: Optional[int] = None,
        return_attention_mask: Optional[bool] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        verbose: bool = True,
    ) -> BatchEncoding:
        """
        Pad a single encoded input or a batch of encoded inputs up to predefined length or to the max sequence length
        in the batch.

        Padding side (left/right) padding token ids are defined at the tokenizer level (with `self.padding_side`,
        `self.pad_token_id` and `self.pad_token_type_id`).

        Please note that with a fast tokenizer, using the `__call__` method is faster than using a method to encode the
        text followed by a call to the `pad` method to get a padded encoding.

        <Tip>

        If the `encoded_inputs` passed are dictionary of numpy arrays, PyTorch tensors or TensorFlow tensors, the
        result will use the same type unless you provide a different tensor type with `return_tensors`. In the case of
        PyTorch tensors, you will lose the specific device of your tensors however.

        </Tip>

        Args:
            encoded_inputs ([`BatchEncoding`], list of [`BatchEncoding`], `Dict[str, List[int]]`, `Dict[str, List[List[int]]` or `List[Dict[str, List[int]]]`):
                Tokenized inputs. Can represent one input ([`BatchEncoding`] or `Dict[str, List[int]]`) or a batch of
                tokenized inputs (list of [`BatchEncoding`], *Dict[str, List[List[int]]]* or *List[Dict[str,
                List[int]]]*) so you can use this method during preprocessing as well as in a PyTorch Dataloader
                collate function.

                Instead of `List[int]` you can have tensors (numpy arrays, PyTorch tensors or TensorFlow tensors), see
                the note above for the return type.
            padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `True`):
                 Select a strategy to pad the returned sequences (according to the model's padding side and padding
                 index) among:

                - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
                  sequence if provided).
                - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum
                  acceptable input length for the model if that argument is not provided.
                - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different
                  lengths).
            max_length (`int`, *optional*):
                Maximum length of the returned list and optionally padding length (see above).
            pad_to_multiple_of (`int`, *optional*):
                If set will pad the sequence to a multiple of the provided value.

                This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability
                `>= 7.5` (Volta).
            return_attention_mask (`bool`, *optional*):
                Whether to return the attention mask. If left to the default, will return the attention mask according
                to the specific tokenizer's default, defined by the `return_outputs` attribute.

                [What are attention masks?](../glossary#attention-mask)
            return_tensors (`str` or [`~utils.TensorType`], *optional*):
                If set, will return tensors instead of list of python integers. Acceptable values are:

                - `'tf'`: Return TensorFlow `tf.constant` objects.
                - `'pt'`: Return PyTorch `mindspore.Tensor` objects.
                - `'np'`: Return Numpy `np.ndarray` objects.
            verbose (`bool`, *optional*, defaults to `True`):
                Whether or not to print more information and warnings.
        """
        if self.__class__.__name__.endswith("Fast"):
            if not self.deprecation_warnings.get("Asking-to-pad-a-fast-tokenizer", False):
                logger.warning_advice(
                    f"You're using a {self.__class__.__name__} tokenizer. Please note that with a fast tokenizer,"
                    " using the `__call__` method is faster than using a method to encode the text followed by a call"
                    " to the `pad` method to get a padded encoding."
                )
                self.deprecation_warnings["Asking-to-pad-a-fast-tokenizer"] = True

        # If we have a list of dicts, let's convert it in a dict of lists
        # We do this to allow using this method as a collate_fn function in PyTorch Dataloader
        if isinstance(encoded_inputs, (list, tuple)) and isinstance(encoded_inputs[0], Mapping):
            encoded_inputs = {key: [example[key] for example in encoded_inputs] for key in encoded_inputs[0].keys()}

        # The model's main input name, usually `input_ids`, has be passed for padding
        if self.model_input_names[0] not in encoded_inputs:
            raise ValueError(
                "You should supply an encoding or a list of encodings to this method "
                f"that includes {self.model_input_names[0]}, but you provided {list(encoded_inputs.keys())}"
            )

        required_input = encoded_inputs[self.model_input_names[0]]

        if required_input is None or (isinstance(required_input, Sized) and len(required_input) == 0):
            if return_attention_mask:
                encoded_inputs["attention_mask"] = []
            return encoded_inputs

        # If we have PyTorch/TF/NumPy tensors/arrays as inputs, we cast them as python objects
        # and rebuild them afterwards if no return_tensors is specified
        # Note that we lose the specific device the tensor may be on for PyTorch

        first_element = required_input[0]
        if isinstance(first_element, (list, tuple)):
            # first_element might be an empty list/tuple in some edge cases so we grab the first non empty element.
            for item in required_input:
                if len(item) != 0:
                    first_element = item[0]
                    break
        # At this state, if `first_element` is still a list/tuple, it's an empty one so there is nothing to do.
        if not isinstance(first_element, (int, list, tuple)):
            if isinstance(first_element, mindspore.Tensor):
                return_tensors = "ms" if return_tensors is None else return_tensors
            elif isinstance(first_element, np.ndarray):
                return_tensors = "np" if return_tensors is None else return_tensors
            else:
                raise ValueError(
                    f"type of {first_element} unknown: {type(first_element)}. "
                    "Should be one of a python, numpy, mindsporeobject."
                )

            for key, value in encoded_inputs.items():
                encoded_inputs[key] = to_py_obj(value)

        # Convert padding_strategy in PaddingStrategy
        padding_strategy, _, max_length, _ = self._get_padding_truncation_strategies(
            padding=padding, max_length=max_length, verbose=verbose
        )

        required_input = encoded_inputs[self.model_input_names[0]]
        if required_input and not isinstance(required_input[0], (list, tuple)):
            encoded_inputs = self._pad(
                encoded_inputs,
                max_length=max_length,
                padding_strategy=padding_strategy,
                pad_to_multiple_of=pad_to_multiple_of,
                return_attention_mask=return_attention_mask,
            )
            return BatchEncoding(encoded_inputs, tensor_type=return_tensors)

        batch_size = len(required_input)
        assert all(
            len(v) == batch_size for v in encoded_inputs.values()
        ), "Some items in the output dictionary have a different batch size than others."

        if padding_strategy == PaddingStrategy.LONGEST:
            max_length = max(len(inputs) for inputs in required_input)
            padding_strategy = PaddingStrategy.MAX_LENGTH

        batch_outputs = {}
        for i in range(batch_size):
            inputs = {k: v[i] for k, v in encoded_inputs.items()}
            outputs = self._pad(
                inputs,
                max_length=max_length,
                padding_strategy=padding_strategy,
                pad_to_multiple_of=pad_to_multiple_of,
                return_attention_mask=return_attention_mask,
            )

            for key, value in outputs.items():
                if key not in batch_outputs:
                    batch_outputs[key] = []
                batch_outputs[key].append(value)

        return BatchEncoding(batch_outputs, tensor_type=return_tensors)

    def create_token_type_ids_from_sequences(
        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
    ) -> List[int]:
        """
        Create the token type IDs corresponding to the sequences passed. [What are token type
        IDs?](../glossary#token-type-ids)

        Should be overridden in a subclass if the model has a special way of building those.

        Args:
            token_ids_0 (`List[int]`): The first tokenized sequence.
            token_ids_1 (`List[int]`, *optional*): The second tokenized sequence.

        Returns:
            `List[int]`: The token type ids.
        """
        if token_ids_1 is None:
            return len(token_ids_0) * [0]
        return [0] * len(token_ids_0) + [1] * len(token_ids_1)

    def build_inputs_with_special_tokens(
        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
    ) -> List[int]:
        """
        Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and
        adding special tokens.

        This implementation does not add special tokens and this method should be overridden in a subclass.

        Args:
            token_ids_0 (`List[int]`): The first tokenized sequence.
            token_ids_1 (`List[int]`, *optional*): The second tokenized sequence.

        Returns:
            `List[int]`: The model input with special tokens.
        """
        if token_ids_1 is None:
            return token_ids_0
        return token_ids_0 + token_ids_1

    def prepare_for_model(
        self,
        ids: List[int],
        pair_ids: Optional[List[int]] = None,
        add_special_tokens: bool = True,
        padding: Union[bool, str, PaddingStrategy] = False,
        truncation: Union[bool, str, TruncationStrategy] = None,
        max_length: Optional[int] = None,
        stride: int = 0,
        pad_to_multiple_of: Optional[int] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        return_token_type_ids: Optional[bool] = None,
        return_attention_mask: Optional[bool] = None,
        return_overflowing_tokens: bool = False,
        return_special_tokens_mask: bool = False,
        return_offsets_mapping: bool = False,
        return_length: bool = False,
        verbose: bool = True,
        prepend_batch_axis: bool = False,
        **kwargs,
    ) -> BatchEncoding:
        """
        Prepares a sequence of input id, or a pair of sequences of inputs ids so that it can be used by the model. It
        adds special tokens, truncates sequences if overflowing while taking into account the special tokens and
        manages a moving window (with user defined stride) for overflowing tokens. Please Note, for *pair_ids*
        different than `None` and *truncation_strategy = longest_first* or `True`, it is not possible to return
        overflowing tokens. Such a combination of arguments will raise an error.

        Args:
            ids (`List[int]`):
                Tokenized input ids of the first sequence. Can be obtained from a string by chaining the `tokenize` and
                `convert_tokens_to_ids` methods.
            pair_ids (`List[int]`, *optional*):
                Tokenized input ids of the second sequence. Can be obtained from a string by chaining the `tokenize`
                and `convert_tokens_to_ids` methods.
        """

        # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
        padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            pad_to_multiple_of=pad_to_multiple_of,
            verbose=verbose,
            **kwargs,
        )

        pair = bool(pair_ids is not None)
        len_ids = len(ids)
        len_pair_ids = len(pair_ids) if pair else 0

        if return_token_type_ids and not add_special_tokens:
            raise ValueError(
                "Asking to return token_type_ids while setting add_special_tokens to False "
                "results in an undefined behavior. Please set add_special_tokens to True or "
                "set return_token_type_ids to None."
            )

        if (
            return_overflowing_tokens
            and truncation_strategy == TruncationStrategy.LONGEST_FIRST
            and pair_ids is not None
        ):
            raise ValueError(
                "Not possible to return overflowing tokens for pair of sequences with the "
                "`longest_first`. Please select another truncation strategy than `longest_first`, "
                "for instance `only_second` or `only_first`."
            )

        # Load from model defaults
        if return_token_type_ids is None:
            return_token_type_ids = "token_type_ids" in self.model_input_names
        if return_attention_mask is None:
            return_attention_mask = "attention_mask" in self.model_input_names

        encoded_inputs = {}

        # Compute the total size of the returned encodings
        total_len = len_ids + len_pair_ids + (self.num_special_tokens_to_add(pair=pair) if add_special_tokens else 0)

        # Truncation: Handle max sequence length
        overflowing_tokens = []
        if truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE and max_length and total_len > max_length:
            ids, pair_ids, overflowing_tokens = self.truncate_sequences(
                ids,
                pair_ids=pair_ids,
                num_tokens_to_remove=total_len - max_length,
                truncation_strategy=truncation_strategy,
                stride=stride,
            )

        if return_overflowing_tokens:
            encoded_inputs["overflowing_tokens"] = overflowing_tokens
            encoded_inputs["num_truncated_tokens"] = total_len - max_length

        # Add special tokens
        if add_special_tokens:
            sequence = self.build_inputs_with_special_tokens(ids, pair_ids)
            token_type_ids = self.create_token_type_ids_from_sequences(ids, pair_ids)
        else:
            sequence = ids + pair_ids if pair else ids
            token_type_ids = [0] * len(ids) + ([0] * len(pair_ids) if pair else [])

        # Build output dictionary
        encoded_inputs["input_ids"] = sequence
        if return_token_type_ids:
            encoded_inputs["token_type_ids"] = token_type_ids
        if return_special_tokens_mask:
            if add_special_tokens:
                encoded_inputs["special_tokens_mask"] = self.get_special_tokens_mask(ids, pair_ids)
            else:
                encoded_inputs["special_tokens_mask"] = [0] * len(sequence)

        # Check lengths
        self._eventual_warn_about_too_long_sequence(encoded_inputs["input_ids"], max_length, verbose)

        # Padding
        if padding_strategy != PaddingStrategy.DO_NOT_PAD or return_attention_mask:
            encoded_inputs = self.pad(
                encoded_inputs,
                max_length=max_length,
                padding=padding_strategy.value,
                pad_to_multiple_of=pad_to_multiple_of,
                return_attention_mask=return_attention_mask,
            )

        if return_length:
            encoded_inputs["length"] = len(encoded_inputs["input_ids"])

        batch_outputs = BatchEncoding(
            encoded_inputs, tensor_type=return_tensors, prepend_batch_axis=prepend_batch_axis
        )

        return batch_outputs

    def truncate_sequences(
        self,
        ids: List[int],
        pair_ids: Optional[List[int]] = None,
        num_tokens_to_remove: int = 0,
        truncation_strategy: Union[str, TruncationStrategy] = "longest_first",
        stride: int = 0,
    ) -> Tuple[List[int], List[int], List[int]]:
        """
        Truncates a sequence pair in-place following the strategy.

        Args:
            ids (`List[int]`):
                Tokenized input ids of the first sequence. Can be obtained from a string by chaining the `tokenize` and
                `convert_tokens_to_ids` methods.
            pair_ids (`List[int]`, *optional*):
                Tokenized input ids of the second sequence. Can be obtained from a string by chaining the `tokenize`
                and `convert_tokens_to_ids` methods.
            num_tokens_to_remove (`int`, *optional*, defaults to 0):
                Number of tokens to remove using the truncation strategy.
            truncation_strategy (`str` or [`~tokenization_utils_base.TruncationStrategy`], *optional*, defaults to `'longest_first'`):
                The strategy to follow for truncation. Can be:

                - `'longest_first'`: Truncate to a maximum length specified with the argument `max_length` or to the
                  maximum acceptable input length for the model if that argument is not provided. This will truncate
                  token by token, removing a token from the longest sequence in the pair if a pair of sequences (or a
                  batch of pairs) is provided.
                - `'only_first'`: Truncate to a maximum length specified with the argument `max_length` or to the
                  maximum acceptable input length for the model if that argument is not provided. This will only
                  truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
                - `'only_second'`: Truncate to a maximum length specified with the argument `max_length` or to the
                  maximum acceptable input length for the model if that argument is not provided. This will only
                  truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
                - `'do_not_truncate'` (default): No truncation (i.e., can output batch with sequence lengths greater
                  than the model maximum admissible input size).
            stride (`int`, *optional*, defaults to 0):
                If set to a positive number, the overflowing tokens returned will contain some tokens from the main
                sequence returned. The value of this argument defines the number of additional tokens.

        Returns:
            `Tuple[List[int], List[int], List[int]]`: The truncated `ids`, the truncated `pair_ids` and the list of
            overflowing tokens. Note: The *longest_first* strategy returns empty list of overflowing tokens if a pair
            of sequences (or a batch of pairs) is provided.
        """
        if num_tokens_to_remove <= 0:
            return ids, pair_ids, []

        if not isinstance(truncation_strategy, TruncationStrategy):
            truncation_strategy = TruncationStrategy(truncation_strategy)

        overflowing_tokens = []
        if truncation_strategy == TruncationStrategy.ONLY_FIRST or (
            truncation_strategy == TruncationStrategy.LONGEST_FIRST and pair_ids is None
        ):
            if len(ids) > num_tokens_to_remove:
                window_len = min(len(ids), stride + num_tokens_to_remove)
                if self.truncation_side == "left":
                    overflowing_tokens = ids[:window_len]
                    ids = ids[num_tokens_to_remove:]
                elif self.truncation_side == "right":
                    overflowing_tokens = ids[-window_len:]
                    ids = ids[:-num_tokens_to_remove]
                else:
                    raise ValueError(f"invalid truncation strategy: {self.truncation_side}, use 'left' or 'right'.")

            else:
                error_msg = (
                    f"We need to remove {num_tokens_to_remove} to truncate the input "
                    f"but the first sequence has a length {len(ids)}. "
                )
                if truncation_strategy == TruncationStrategy.ONLY_FIRST:
                    error_msg = (
                        error_msg + "Please select another truncation strategy than "
                        f"{truncation_strategy}, for instance 'longest_first' or 'only_second'."
                    )
                logger.error(error_msg)
        elif truncation_strategy == TruncationStrategy.LONGEST_FIRST:
            logger.warning(
                "Be aware, overflowing tokens are not returned for the setting you have chosen,"
                f" i.e. sequence pairs with the '{TruncationStrategy.LONGEST_FIRST.value}' "
                "truncation strategy. So the returned list will always be empty even if some "
                "tokens have been removed."
            )
            len_pair_ids = len(pair_ids) if pair_ids is not None else 0
            len_ids = len(ids)
            first_remove = min(abs(len_pair_ids - len_ids), num_tokens_to_remove)
            second_remove = num_tokens_to_remove - first_remove
            if len_ids > len_pair_ids:
                ids_to_move = first_remove + second_remove // 2
                pair_ids_to_move = second_remove - second_remove // 2
            else:
                ids_to_move = second_remove // 2
                pair_ids_to_move = first_remove + second_remove - (second_remove // 2)

            if self.truncation_side == "right":
                ids = ids[:-ids_to_move] if ids_to_move > 0 else ids
                pair_ids = pair_ids[:-pair_ids_to_move] if pair_ids is not None and pair_ids_to_move > 0 else pair_ids
            elif self.truncation_side == "left":
                ids = ids[ids_to_move:]
                pair_ids = pair_ids[pair_ids_to_move:] if pair_ids is not None else None
            else:
                raise ValueError(f"invalid truncation strategy:{self.truncation_side}")

        elif truncation_strategy == TruncationStrategy.ONLY_SECOND and pair_ids is not None:
            if len(pair_ids) > num_tokens_to_remove:
                window_len = min(len(pair_ids), stride + num_tokens_to_remove)
                if self.truncation_side == "right":
                    overflowing_tokens = pair_ids[-window_len:]
                    pair_ids = pair_ids[:-num_tokens_to_remove]
                elif self.truncation_side == "left":
                    overflowing_tokens = pair_ids[:window_len]
                    pair_ids = pair_ids[num_tokens_to_remove:]
                else:
                    raise ValueError(f"invalid truncation strategy:{self.truncation_side}")
            else:
                logger.error(
                    f"We need to remove {num_tokens_to_remove} to truncate the input "
                    f"but the second sequence has a length {len(pair_ids)}. "
                    f"Please select another truncation strategy than {truncation_strategy}, "
                    "for instance 'longest_first' or 'only_first'."
                )

        return (ids, pair_ids, overflowing_tokens)

    def _pad(
        self,
        encoded_inputs: Union[Dict[str, EncodedInput], BatchEncoding],
        max_length: Optional[int] = None,
        padding_strategy: PaddingStrategy = PaddingStrategy.DO_NOT_PAD,
        pad_to_multiple_of: Optional[int] = None,
        return_attention_mask: Optional[bool] = None,
    ) -> dict:
        """
        Pad encoded inputs (on left/right and up to predefined length or max length in the batch)

        Args:
            encoded_inputs:
                Dictionary of tokenized inputs (`List[int]`) or batch of tokenized inputs (`List[List[int]]`).
            max_length: maximum length of the returned list and optionally padding length (see below).
                Will truncate by taking into account the special tokens.
            padding_strategy: PaddingStrategy to use for padding.

                - PaddingStrategy.LONGEST Pad to the longest sequence in the batch
                - PaddingStrategy.MAX_LENGTH: Pad to the max length (default)
                - PaddingStrategy.DO_NOT_PAD: Do not pad
                The tokenizer padding sides are defined in self.padding_side:

                    - 'left': pads on the left of the sequences
                    - 'right': pads on the right of the sequences
            pad_to_multiple_of: (optional) Integer if set will pad the sequence to a multiple of the provided value.
                This is especially useful to enable the use of Tensor Core on NVIDIA hardware with compute capability
                `>= 7.5` (Volta).
            return_attention_mask:
                (optional) Set to False to avoid returning attention mask (default: set to model specifics)
        """
        # Load from model defaults
        if return_attention_mask is None:
            return_attention_mask = "attention_mask" in self.model_input_names

        required_input = encoded_inputs[self.model_input_names[0]]

        if padding_strategy == PaddingStrategy.LONGEST:
            max_length = len(required_input)

        if max_length is not None and pad_to_multiple_of is not None and (max_length % pad_to_multiple_of != 0):
            max_length = ((max_length // pad_to_multiple_of) + 1) * pad_to_multiple_of

        needs_to_be_padded = padding_strategy != PaddingStrategy.DO_NOT_PAD and len(required_input) != max_length

        # Initialize attention mask if not present.
        if return_attention_mask and "attention_mask" not in encoded_inputs:
            encoded_inputs["attention_mask"] = [1] * len(required_input)

        if needs_to_be_padded:
            difference = max_length - len(required_input)

            if self.padding_side == "right":
                if return_attention_mask:
                    encoded_inputs["attention_mask"] = encoded_inputs["attention_mask"] + [0] * difference
                if "token_type_ids" in encoded_inputs:
                    encoded_inputs["token_type_ids"] = (
                        encoded_inputs["token_type_ids"] + [self.pad_token_type_id] * difference
                    )
                if "special_tokens_mask" in encoded_inputs:
                    encoded_inputs["special_tokens_mask"] = encoded_inputs["special_tokens_mask"] + [1] * difference
                encoded_inputs[self.model_input_names[0]] = required_input + [self.pad_token_id] * difference
            elif self.padding_side == "left":
                if return_attention_mask:
                    encoded_inputs["attention_mask"] = [0] * difference + encoded_inputs["attention_mask"]
                if "token_type_ids" in encoded_inputs:
                    encoded_inputs["token_type_ids"] = [self.pad_token_type_id] * difference + encoded_inputs[
                        "token_type_ids"
                    ]
                if "special_tokens_mask" in encoded_inputs:
                    encoded_inputs["special_tokens_mask"] = [1] * difference + encoded_inputs["special_tokens_mask"]
                encoded_inputs[self.model_input_names[0]] = [self.pad_token_id] * difference + required_input
            else:
                raise ValueError(f"Invalid padding strategy:{self.padding_side}")

        return encoded_inputs

    def convert_tokens_to_string(self, tokens: List[str]) -> str:
        """
        Converts a sequence of tokens in a single string. The most simple way to do it is `" ".join(tokens)` but we
        often want to remove sub-word tokenization artifacts at the same time.

        Args:
            tokens (`List[str]`): The token to join in a string.

        Returns:
            `str`: The joined tokens.
        """
        raise NotImplementedError

    def batch_decode(
        self,
        sequences: Union[List[int], List[List[int]], "np.ndarray", "mindspore.Tensor"],
        skip_special_tokens: bool = False,
        clean_up_tokenization_spaces: bool = None,
        **kwargs,
    ) -> List[str]:
        """
        Convert a list of lists of token ids into a list of strings by calling decode.

        Args:
            sequences (`Union[List[int], List[List[int]], np.ndarray, mindspore.Tensor, tf.Tensor]`):
                List of tokenized input ids. Can be obtained using the `__call__` method.
            skip_special_tokens (`bool`, *optional*, defaults to `False`):
                Whether or not to remove special tokens in the decoding.
            clean_up_tokenization_spaces (`bool`, *optional*):
                Whether or not to clean up the tokenization spaces. If `None`, will default to
                `self.clean_up_tokenization_spaces`.
            kwargs (additional keyword arguments, *optional*):
                Will be passed to the underlying model specific decode method.

        Returns:
            `List[str]`: The list of decoded sentences.
        """
        return [
            self.decode(
                seq,
                skip_special_tokens=skip_special_tokens,
                clean_up_tokenization_spaces=clean_up_tokenization_spaces,
                **kwargs,
            )
            for seq in sequences
        ]

    def decode(
        self,
        token_ids: Union[int, List[int], "np.ndarray", "mindspore.Tensor"],
        skip_special_tokens: bool = False,
        clean_up_tokenization_spaces: bool = None,
        **kwargs,
    ) -> str:
        """
        Converts a sequence of ids in a string, using the tokenizer and vocabulary with options to remove special
        tokens and clean up tokenization spaces.

        Similar to doing `self.convert_tokens_to_string(self.convert_ids_to_tokens(token_ids))`.

        Args:
            token_ids (`Union[int, List[int], np.ndarray, mindspore.Tensor, tf.Tensor]`):
                List of tokenized input ids. Can be obtained using the `__call__` method.
            skip_special_tokens (`bool`, *optional*, defaults to `False`):
                Whether or not to remove special tokens in the decoding.
            clean_up_tokenization_spaces (`bool`, *optional*):
                Whether or not to clean up the tokenization spaces. If `None`, will default to
                `self.clean_up_tokenization_spaces`.
            kwargs (additional keyword arguments, *optional*):
                Will be passed to the underlying model specific decode method.

        Returns:
            `str`: The decoded sentence.
        """
        # Convert inputs to python lists
        token_ids = to_py_obj(token_ids)

        return self._decode(
            token_ids=token_ids,
            skip_special_tokens=skip_special_tokens,
            clean_up_tokenization_spaces=clean_up_tokenization_spaces,
            **kwargs,
        )

    def _decode(
        self,
        token_ids: Union[int, List[int]],
        skip_special_tokens: bool = False,
        clean_up_tokenization_spaces: bool = None,
        **kwargs,
    ) -> str:
        raise NotImplementedError

    def get_special_tokens_mask(
        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
    ) -> List[int]:
        """
        Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding
        special tokens using the tokenizer `prepare_for_model` or `encode_plus` methods.

        Args:
            token_ids_0 (`List[int]`):
                List of ids of the first sequence.
            token_ids_1 (`List[int]`, *optional*):
                List of ids of the second sequence.
            already_has_special_tokens (`bool`, *optional*, defaults to `False`):
                Whether or not the token list is already formatted with special tokens for the model.

        Returns:
            A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
        """
        assert already_has_special_tokens and token_ids_1 is None, (
            "You cannot use ``already_has_special_tokens=False`` with this tokenizer. "
            "Please use a slow (full python) tokenizer to activate this argument. "
            "Or set `return_special_tokens_mask=True` when calling the encoding method "
            "to get the special tokens mask in any tokenizer. "
        )

        all_special_ids = self.all_special_ids  # cache the property

        special_tokens_mask = [1 if token in all_special_ids else 0 for token in token_ids_0]

        return special_tokens_mask

    @staticmethod
    def clean_up_tokenization(out_string: str) -> str:
        """
        Clean up a list of simple English tokenization artifacts like spaces before punctuations and abbreviated forms.

        Args:
            out_string (`str`): The text to clean up.

        Returns:
            `str`: The cleaned-up string.
        """
        out_string = (
            out_string.replace(" .", ".")
            .replace(" ?", "?")
            .replace(" !", "!")
            .replace(" ,", ",")
            .replace(" ' ", "'")
            .replace(" n't", "n't")
            .replace(" 'm", "'m")
            .replace(" 's", "'s")
            .replace(" 've", "'ve")
            .replace(" 're", "'re")
        )
        return out_string

    def _eventual_warn_about_too_long_sequence(self, ids: List[int], max_length: Optional[int], verbose: bool):
        """
        Depending on the input and internal state we might trigger a warning about a sequence that is too long for its
        corresponding model

        Args:
            ids (`List[str]`): The ids produced by the tokenization
            max_length (`int`, *optional*): The max_length desired (does not trigger a warning if it is set)
            verbose (`bool`): Whether or not to print more information and warnings.

        """
        if max_length is None and len(ids) > self.model_max_length and verbose:
            if not self.deprecation_warnings.get("sequence-length-is-longer-than-the-specified-maximum", False):
                logger.warning(
                    "Token indices sequence length is longer than the specified maximum sequence length "
                    f"for this model ({len(ids)} > {self.model_max_length}). Running this sequence through the model "
                    "will result in indexing errors"
                )
            self.deprecation_warnings["sequence-length-is-longer-than-the-specified-maximum"] = True

    def _switch_to_input_mode(self):
        """
        Private method to put the tokenizer in input mode (when it has different modes for input/outputs)
        """

    def _switch_to_target_mode(self):
        """
        Private method to put the tokenizer in target mode (when it has different modes for input/outputs)
        """

    @contextmanager
    def as_target_tokenizer(self):
        """
        Temporarily sets the tokenizer for encoding the targets. Useful for tokenizer associated to
        sequence-to-sequence models that need a slightly different processing for the labels.
        """
        warnings.warn(
            "`as_target_tokenizer` is deprecated. You can tokenize your "
            "labels by using the argument `text_target` of the regular `__call__` method (either in the same call as "
            "your input texts if you use the same keyword arguments, or in a separate call."
        )
        self._switch_to_target_mode()
        self._in_target_context_manager = True
        yield
        self._in_target_context_manager = False
        self._switch_to_input_mode()

    @classmethod
    def register_for_auto_class(cls, auto_class="AutoTokenizer"):
        """
        Register this class with a given auto class. This should only be used for custom tokenizers as the ones in the
        library are already mapped with `AutoTokenizer`.

        <Tip warning={true}>

        This API is experimental and may have some slight breaking changes in the next releases.

        </Tip>

        Args:
            auto_class (`str` or `type`, *optional*, defaults to `"AutoTokenizer"`):
                The auto class to register this new tokenizer with.
        """
        if not isinstance(auto_class, str):
            auto_class = auto_class.__name__

        import mindnlp.transformers.models.auto as auto_module

        if not hasattr(auto_module, auto_class):
            raise ValueError(f"{auto_class} is not a valid auto class.")

        cls._auto_class = auto_class

    def prepare_seq2seq_batch(
        self,
        src_texts: List[str],
        tgt_texts: Optional[List[str]] = None,
        max_length: Optional[int] = None,
        max_target_length: Optional[int] = None,
        padding: str = "longest",
        return_tensors: str = None,
        truncation: bool = True,
        **kwargs,
    ) -> BatchEncoding:
        """
        Prepare model inputs for translation. For best performance, translate one sentence at a time.

        Arguments:
            src_texts (`List[str]`):
                List of documents to summarize or source language texts.
            tgt_texts (`list`, *optional*):
                List of summaries or target language texts.
            max_length (`int`, *optional*):
                Controls the maximum length for encoder inputs (documents to summarize or source language texts) If
                left unset or set to `None`, this will use the predefined model maximum length if a maximum length is
                required by one of the truncation/padding parameters. If the model has no specific maximum input length
                (like XLNet) truncation/padding to a maximum length will be deactivated.
            max_target_length (`int`, *optional*):
                Controls the maximum length of decoder inputs (target language texts or summaries) If left unset or set
                to `None`, this will use the max_length value.
            padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`):
                Activates and controls padding. Accepts the following values:

                - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
                  sequence if provided).
                - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum
                  acceptable input length for the model if that argument is not provided.
                - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different
                  lengths).
            return_tensors (`str` or [`~utils.TensorType`], *optional*):
                If set, will return tensors instead of list of python integers. Acceptable values are:

                - `'tf'`: Return TensorFlow `tf.constant` objects.
                - `'pt'`: Return PyTorch `mindspore.Tensor` objects.
                - `'np'`: Return Numpy `np.ndarray` objects.
            truncation (`bool`, `str` or [`~tokenization_utils_base.TruncationStrategy`], *optional*, defaults to `True`):
                Activates and controls truncation. Accepts the following values:

                - `True` or `'longest_first'`: Truncate to a maximum length specified with the argument `max_length` or
                  to the maximum acceptable input length for the model if that argument is not provided. This will
                  truncate token by token, removing a token from the longest sequence in the pair if a pair of
                  sequences (or a batch of pairs) is provided.
                - `'only_first'`: Truncate to a maximum length specified with the argument `max_length` or to the
                  maximum acceptable input length for the model if that argument is not provided. This will only
                  truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
                - `'only_second'`: Truncate to a maximum length specified with the argument `max_length` or to the
                  maximum acceptable input length for the model if that argument is not provided. This will only
                  truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
                - `False` or `'do_not_truncate'` (default): No truncation (i.e., can output batch with sequence lengths
                  greater than the model maximum admissible input size).
            **kwargs:
                Additional keyword arguments passed along to `self.__call__`.

        Return:
            [`BatchEncoding`]: A [`BatchEncoding`] with the following fields:

            - **input_ids** -- List of token ids to be fed to the encoder.
            - **attention_mask** -- List of indices specifying which tokens should be attended to by the model.
            - **labels** -- List of token ids for tgt_texts.

            The full set of keys `[input_ids, attention_mask, labels]`, will only be returned if tgt_texts is passed.
            Otherwise, input_ids, attention_mask will be the only keys.
        """
        # docstyle-ignore
        formatted_warning = """
`prepare_seq2seq_batch` is deprecated. Use the regular
`__call__` method to prepare your inputs and targets.

Here is a short example:

model_inputs = tokenizer(src_texts, text_target=tgt_texts, ...)

If you either need to use different keyword arguments for the source and target texts, you should do two calls like
this:

model_inputs = tokenizer(src_texts, ...)
labels = tokenizer(text_target=tgt_texts, ...)
model_inputs["labels"] = labels["input_ids"]

See the documentation of your specific tokenizer for more details on the specific arguments to the tokenizer of choice.
For a more complete example, see the implementation of `prepare_seq2seq_batch`.
"""
        warnings.warn(formatted_warning, FutureWarning)
        # mBART-specific kwargs that should be ignored by other models.
        kwargs.pop("src_lang", None)
        kwargs.pop("tgt_lang", None)
        if max_length is None:
            max_length = self.model_max_length
        model_inputs = self(
            src_texts,
            add_special_tokens=True,
            return_tensors=return_tensors,
            max_length=max_length,
            padding=padding,
            truncation=truncation,
            **kwargs,
        )
        if tgt_texts is None:
            return model_inputs
        # Process tgt_texts
        if max_target_length is None:
            max_target_length = max_length
        with self.as_target_tokenizer():
            labels = self(
                tgt_texts,
                add_special_tokens=True,
                return_tensors=return_tensors,
                padding=padding,
                max_length=max_target_length,
                truncation=truncation,
                **kwargs,
            )
        model_inputs["labels"] = labels["input_ids"]
        return model_inputs

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.max_len_sentences_pair: int property writable

int: The maximum combined length of a pair of sentences that can be fed to the model.

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.max_len_single_sentence: int property writable

int: The maximum length of a sentence that can be fed to the model.

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.__call__(text=None, text_pair=None, text_target=None, text_pair_target=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, return_tensors=None, return_token_type_ids=None, return_attention_mask=None, return_overflowing_tokens=False, return_special_tokens_mask=False, return_offsets_mapping=False, return_length=False, verbose=True, **kwargs)

Main method to tokenize and prepare for the model one or several sequence(s) or one or several pair(s) of sequences.

PARAMETER	DESCRIPTION
text	The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set is_split_into_words=True (to lift the ambiguity with a batch of sequences). TYPE: `str`, `List[str]`, `List[List[str]]`, *optional* DEFAULT: None
text_pair	The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set is_split_into_words=True (to lift the ambiguity with a batch of sequences). TYPE: `str`, `List[str]`, `List[List[str]]`, *optional* DEFAULT: None
text_target	The sequence or batch of sequences to be encoded as target texts. Each sequence can be a string or a list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set is_split_into_words=True (to lift the ambiguity with a batch of sequences). TYPE: `str`, `List[str]`, `List[List[str]]`, *optional* DEFAULT: None
text_pair_target	The sequence or batch of sequences to be encoded as target texts. Each sequence can be a string or a list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set is_split_into_words=True (to lift the ambiguity with a batch of sequences). TYPE: `str`, `List[str]`, `List[List[str]]`, *optional* DEFAULT: None

Source code in mindnlp\transformers\tokenization_utils_base.py

def __call__(
    self,
    text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None,
    text_pair: Optional[Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]]] = None,
    text_target: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None,
    text_pair_target: Optional[
        Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]]
    ] = None,
    add_special_tokens: bool = True,
    padding: Union[bool, str, PaddingStrategy] = False,
    truncation: Union[bool, str, TruncationStrategy] = None,
    max_length: Optional[int] = None,
    stride: int = 0,
    is_split_into_words: bool = False,
    pad_to_multiple_of: Optional[int] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    return_token_type_ids: Optional[bool] = None,
    return_attention_mask: Optional[bool] = None,
    return_overflowing_tokens: bool = False,
    return_special_tokens_mask: bool = False,
    return_offsets_mapping: bool = False,
    return_length: bool = False,
    verbose: bool = True,
    **kwargs,
) -> BatchEncoding:
    """
    Main method to tokenize and prepare for the model one or several sequence(s) or one or several pair(s) of
    sequences.

    Args:
        text (`str`, `List[str]`, `List[List[str]]`, *optional*):
            The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings
            (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set
            `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
        text_pair (`str`, `List[str]`, `List[List[str]]`, *optional*):
            The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings
            (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set
            `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
        text_target (`str`, `List[str]`, `List[List[str]]`, *optional*):
            The sequence or batch of sequences to be encoded as target texts. Each sequence can be a string or a
            list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized),
            you must set `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
        text_pair_target (`str`, `List[str]`, `List[List[str]]`, *optional*):
            The sequence or batch of sequences to be encoded as target texts. Each sequence can be a string or a
            list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized),
            you must set `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
    """
    # To avoid duplicating
    all_kwargs = {
        "add_special_tokens": add_special_tokens,
        "padding": padding,
        "truncation": truncation,
        "max_length": max_length,
        "stride": stride,
        "is_split_into_words": is_split_into_words,
        "pad_to_multiple_of": pad_to_multiple_of,
        "return_tensors": return_tensors,
        "return_token_type_ids": return_token_type_ids,
        "return_attention_mask": return_attention_mask,
        "return_overflowing_tokens": return_overflowing_tokens,
        "return_special_tokens_mask": return_special_tokens_mask,
        "return_offsets_mapping": return_offsets_mapping,
        "return_length": return_length,
        "split_special_tokens": kwargs.pop("split_special_tokens", self.split_special_tokens),
        "verbose": verbose,
    }
    all_kwargs.update(kwargs)
    if text is None and text_target is None:
        raise ValueError("You need to specify either `text` or `text_target`.")
    if text is not None:
        # for mindspore.dataset
        if isinstance(text, np.ndarray):
            text = str(text)
            if isinstance(text_pair, np.ndarray):
                text_pair = str(text_pair)
            elif isinstance(text_pair, list):
                text_pair = [str(t) for t in text_pair]
        # The context manager will send the inputs as normal texts and not text_target, but we shouldn't change the
        # input mode in this case.
        if not self._in_target_context_manager:
            self._switch_to_input_mode()
        encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs)
    if text_target is not None:
        self._switch_to_target_mode()
        target_encodings = self._call_one(text=text_target, text_pair=text_pair_target, **all_kwargs)
    # Leave back tokenizer in input mode
    self._switch_to_input_mode()

    if text_target is None:
        return encodings
    elif text is None:
        return target_encodings
    else:
        encodings["labels"] = target_encodings["input_ids"]
        return encodings

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.apply_chat_template(conversation, tools=None, documents=None, chat_template=None, add_generation_prompt=False, tokenize=True, padding=False, truncation=False, max_length=None, return_tensors=None, return_dict=False, return_assistant_tokens_mask=False, tokenizer_kwargs=None, **kwargs)

Converts a list of dictionaries with "role" and "content" keys to a list of token ids. This method is intended for use with chat models, and will read the tokenizer's chat_template attribute to determine the format and control tokens to use when converting.

PARAMETER	DESCRIPTION
conversation	A list of dicts with "role" and "content" keys, representing the chat history so far. TYPE: Union[List[Dict[str, str]], List[List[Dict[str, str]]]]
tools	A list of tools (callable functions) that will be accessible to the model. If the template does not support function calling, this argument will have no effect. Each tool should be passed as a JSON Schema, giving the name, description and argument types for the tool. See our chat templating guide for more information. TYPE: `List[Dict]`, *optional* DEFAULT: None
documents	A list of dicts representing documents that will be accessible to the model if it is performing RAG (retrieval-augmented generation). If the template does not support RAG, this argument will have no effect. We recommend that each document should be a dict containing "title" and "text" keys. Please see the RAG section of the chat templating guide for examples of passing documents with chat templates. TYPE: `List[Dict[str, str]]`, *optional* DEFAULT: None
chat_template	A Jinja template to use for this conversion. It is usually not necessary to pass anything to this argument, as the model's template will be used by default. TYPE: `str`, *optional* DEFAULT: None
add_generation_prompt	Whether to end the prompt with the token(s) that indicate the start of an assistant message. This is useful when you want to generate a response from the model. Note that this argument will be passed to the chat template, and so it must be supported in the template for this argument to have any effect. TYPE: bool, *optional* DEFAULT: False
tokenize	Whether to tokenize the output. If False, the output will be a string. TYPE: `bool`, defaults to `True` DEFAULT: True
padding	Whether to pad sequences to the maximum length. Has no effect if tokenize is False. TYPE: `bool`, defaults to `False` DEFAULT: False
truncation	Whether to truncate sequences at the maximum length. Has no effect if tokenize is False. TYPE: `bool`, defaults to `False` DEFAULT: False
max_length	Maximum length (in tokens) to use for padding or truncation. Has no effect if tokenize is False. If not specified, the tokenizer's max_length attribute will be used as a default. TYPE: `int`, *optional* DEFAULT: None
return_tensors	If set, will return tensors of a particular framework. Has no effect if tokenize is False. Acceptable values are: - 'tf': Return TensorFlow tf.Tensor objects. - 'pt': Return PyTorch mindspore.Tensor objects. - 'np': Return NumPy np.ndarray objects. - 'jax': Return JAX jnp.ndarray objects. TYPE: `str` or [`~utils.TensorType`], *optional* DEFAULT: None
return_dict	Whether to return a dictionary with named outputs. Has no effect if tokenize is False. TYPE: `bool`, defaults to `False` DEFAULT: False
tokenizer_kwargs	Any]`, optional): Additional kwargs to pass to the tokenizer. TYPE: `Dict[str DEFAULT: None
return_assistant_tokens_mask	Whether to return a mask of the assistant generated tokens. For tokens generated by the assistant, the mask will contain 1. For user and system tokens, the mask will contain 0. This functionality is only available for chat templates that support it via the {% generation %} keyword. TYPE: `bool`, defaults to `False` DEFAULT: False
**kwargs	Additional kwargs to pass to the template renderer. Will be accessible by the chat template. DEFAULT: {}

RETURNS	DESCRIPTION
Union[str, List[int], List[str], List[List[int]], BatchEncoding]	Union[List[int], Dict]: A list of token ids representing the tokenized chat so far, including control tokens. This
Union[str, List[int], List[str], List[List[int]], BatchEncoding]	output is ready to pass to the model, either directly or via methods like generate(). If return_dict is
Union[str, List[int], List[str], List[List[int]], BatchEncoding]	set, will return a dict of tokenizer outputs instead.

Source code in mindnlp\transformers\tokenization_utils_base.py

def apply_chat_template(
    self,
    conversation: Union[List[Dict[str, str]], List[List[Dict[str, str]]]],
    tools: Optional[List[Dict]] = None,
    documents: Optional[List[Dict[str, str]]] = None,
    chat_template: Optional[str] = None,
    add_generation_prompt: bool = False,
    tokenize: bool = True,
    padding: bool = False,
    truncation: bool = False,
    max_length: Optional[int] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    return_dict: bool = False,
    return_assistant_tokens_mask: bool = False,
    tokenizer_kwargs: Optional[Dict[str, Any]] = None,
    **kwargs,
) -> Union[str, List[int], List[str], List[List[int]], BatchEncoding]:
    """
    Converts a list of dictionaries with `"role"` and `"content"` keys to a list of token
    ids. This method is intended for use with chat models, and will read the tokenizer's chat_template attribute to
    determine the format and control tokens to use when converting.

    Args:
        conversation (Union[List[Dict[str, str]], List[List[Dict[str, str]]]]): A list of dicts
            with "role" and "content" keys, representing the chat history so far.
        tools (`List[Dict]`, *optional*):
            A list of tools (callable functions) that will be accessible to the model. If the template does not
            support function calling, this argument will have no effect. Each tool should be passed as a JSON Schema,
            giving the name, description and argument types for the tool. See our
            [chat templating guide](https://huggingface.co/docs/transformers/main/en/chat_templating#automated-function-conversion-for-tool-use)
            for more information.
        documents (`List[Dict[str, str]]`, *optional*):
            A list of dicts representing documents that will be accessible to the model if it is performing RAG
            (retrieval-augmented generation). If the template does not support RAG, this argument will have no
            effect. We recommend that each document should be a dict containing "title" and "text" keys. Please
            see the RAG section of the [chat templating guide](https://huggingface.co/docs/transformers/main/en/chat_templating#arguments-for-RAG)
            for examples of passing documents with chat templates.
        chat_template (`str`, *optional*):
            A Jinja template to use for this conversion. It is usually not necessary to pass anything to this
            argument, as the model's template will be used by default.
        add_generation_prompt (bool, *optional*): Whether to end the prompt with the token(s) that indicate
            the start of an assistant message. This is useful when you want to generate a response from the model.
            Note that this argument will be passed to the chat template, and so it must be supported in the
            template for this argument to have any effect.
        tokenize (`bool`, defaults to `True`):
            Whether to tokenize the output. If `False`, the output will be a string.
        padding (`bool`, defaults to `False`):
            Whether to pad sequences to the maximum length. Has no effect if tokenize is `False`.
        truncation (`bool`, defaults to `False`):
            Whether to truncate sequences at the maximum length. Has no effect if tokenize is `False`.
        max_length (`int`, *optional*):
            Maximum length (in tokens) to use for padding or truncation. Has no effect if tokenize is `False`. If
            not specified, the tokenizer's `max_length` attribute will be used as a default.
        return_tensors (`str` or [`~utils.TensorType`], *optional*):
            If set, will return tensors of a particular framework. Has no effect if tokenize is `False`. Acceptable
            values are:
            - `'tf'`: Return TensorFlow `tf.Tensor` objects.
            - `'pt'`: Return PyTorch `mindspore.Tensor` objects.
            - `'np'`: Return NumPy `np.ndarray` objects.
            - `'jax'`: Return JAX `jnp.ndarray` objects.
        return_dict (`bool`, defaults to `False`):
            Whether to return a dictionary with named outputs. Has no effect if tokenize is `False`.
        tokenizer_kwargs (`Dict[str: Any]`, *optional*): Additional kwargs to pass to the tokenizer.
        return_assistant_tokens_mask (`bool`, defaults to `False`):
            Whether to return a mask of the assistant generated tokens. For tokens generated by the assistant,
            the mask will contain 1. For user and system tokens, the mask will contain 0.
            This functionality is only available for chat templates that support it via the `{% generation %}` keyword.
        **kwargs: Additional kwargs to pass to the template renderer. Will be accessible by the chat template.

    Returns:
        `Union[List[int], Dict]`: A list of token ids representing the tokenized chat so far, including control tokens. This
        output is ready to pass to the model, either directly or via methods like `generate()`. If `return_dict` is
        set, will return a dict of tokenizer outputs instead.
    """

    if return_dict and not tokenize:
        raise ValueError(
            "`return_dict=True` is incompatible with `tokenize=False`, because there is no dict "
            "of tokenizer outputs to return."
        )

    if return_assistant_tokens_mask and not return_dict:
        raise ValueError("`return_assistant_tokens_mask=True` is incompatible with `return_dict=False`")

    if tokenizer_kwargs is None:
        tokenizer_kwargs = {}

    chat_template = self.get_chat_template(chat_template, tools)

    if return_assistant_tokens_mask and not re.search(r"\{\%-?\s*generation\s*-?\%\}", chat_template):
        logger.warning_once(
            "return_assistant_tokens_mask==True but chat template does not contain `{% generation %}` keyword."
        )

    # Compilation function uses a cache to avoid recompiling the same template
    compiled_template = self._compile_jinja_template(chat_template)

    if isinstance(conversation, (list, tuple)) and (
        isinstance(conversation[0], (list, tuple)) or hasattr(conversation[0], "messages")
    ):
        conversations = conversation
        is_batched = True
    else:
        conversations = [conversation]
        is_batched = False

    # We accept either JSON schemas or functions for tools. If we get functions, we convert them to schemas
    if tools is not None:
        tool_schemas = []
        for tool in tools:
            if isinstance(tool, dict):
                tool_schemas.append(tool)
            elif isfunction(tool):
                tool_schemas.append(get_json_schema(tool))
            else:
                raise ValueError(
                    "Tools should either be a JSON schema, or a callable function with type hints "
                    "and a docstring suitable for auto-conversion to a schema."
                )
    else:
        tool_schemas = None

    if documents is not None:
        for document in documents:
            if not isinstance(document, dict):
                raise TypeError("Documents should be a list of dicts with 'title' and 'text' keys!")

    rendered = []
    all_generation_indices = []
    template_kwargs = {**self.special_tokens_map, **kwargs}  # kwargs overwrite special tokens if both are present
    for chat in conversations:
        if hasattr(chat, "messages"):
            # Indicates it's a Conversation object
            chat = chat.messages
        if return_assistant_tokens_mask:
            rendered_chat, generation_indices = self._render_with_assistant_indices(
                compiled_template=compiled_template,
                messages=chat,
                tools=tool_schemas,
                documents=documents,
                add_generation_prompt=add_generation_prompt,
                **template_kwargs,
            )
            all_generation_indices.append(generation_indices)
        else:
            rendered_chat = compiled_template.render(
                messages=chat,
                tools=tool_schemas,
                documents=documents,
                add_generation_prompt=add_generation_prompt,
                **template_kwargs,
            )
        rendered.append(rendered_chat)

    if not is_batched:
        rendered = rendered[0]

    if tokenize:
        out = self(
            rendered,
            padding=padding,
            truncation=truncation,
            max_length=max_length,
            add_special_tokens=False,
            return_tensors=return_tensors,
            **tokenizer_kwargs,
        )
        if return_dict:
            if return_assistant_tokens_mask:
                assistant_masks = []
                if is_batched or return_tensors:
                    input_ids = out["input_ids"]
                else:
                    input_ids = [out["input_ids"]]
                for i in range(len(input_ids)):
                    current_mask = [0] * len(input_ids[i])
                    for assistant_start_char, assistant_end_char in all_generation_indices[i]:
                        start_token = out.char_to_token(i, assistant_start_char)
                        end_token = out.char_to_token(i, assistant_end_char - 1)
                        if start_token is None:
                            # start_token is out of bounds maybe due to truncation.
                            break
                        for token_id in range(start_token, end_token + 1 if end_token else len(input_ids)):
                            current_mask[token_id] = 1
                    assistant_masks.append(current_mask)
                out["assistant_masks"] = assistant_masks if is_batched else assistant_masks[0]
            return out
        else:
            return out["input_ids"]
    else:
        return rendered

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.as_target_tokenizer()

Temporarily sets the tokenizer for encoding the targets. Useful for tokenizer associated to sequence-to-sequence models that need a slightly different processing for the labels.

Source code in mindnlp\transformers\tokenization_utils_base.py

@contextmanager
def as_target_tokenizer(self):
    """
    Temporarily sets the tokenizer for encoding the targets. Useful for tokenizer associated to
    sequence-to-sequence models that need a slightly different processing for the labels.
    """
    warnings.warn(
        "`as_target_tokenizer` is deprecated. You can tokenize your "
        "labels by using the argument `text_target` of the regular `__call__` method (either in the same call as "
        "your input texts if you use the same keyword arguments, or in a separate call."
    )
    self._switch_to_target_mode()
    self._in_target_context_manager = True
    yield
    self._in_target_context_manager = False
    self._switch_to_input_mode()

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.batch_decode(sequences, skip_special_tokens=False, clean_up_tokenization_spaces=None, **kwargs)

Convert a list of lists of token ids into a list of strings by calling decode.

PARAMETER	DESCRIPTION
sequences	List of tokenized input ids. Can be obtained using the __call__ method. TYPE: `Union[List[int], List[List[int]], np.ndarray, mindspore.Tensor, tf.Tensor]`
skip_special_tokens	Whether or not to remove special tokens in the decoding. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False
clean_up_tokenization_spaces	Whether or not to clean up the tokenization spaces. If None, will default to self.clean_up_tokenization_spaces. TYPE: `bool`, *optional* DEFAULT: None
kwargs	Will be passed to the underlying model specific decode method. TYPE: additional keyword arguments, *optional* DEFAULT: {}

RETURNS	DESCRIPTION
List[str]	List[str]: The list of decoded sentences.

Source code in mindnlp\transformers\tokenization_utils_base.py

def batch_decode(
    self,
    sequences: Union[List[int], List[List[int]], "np.ndarray", "mindspore.Tensor"],
    skip_special_tokens: bool = False,
    clean_up_tokenization_spaces: bool = None,
    **kwargs,
) -> List[str]:
    """
    Convert a list of lists of token ids into a list of strings by calling decode.

    Args:
        sequences (`Union[List[int], List[List[int]], np.ndarray, mindspore.Tensor, tf.Tensor]`):
            List of tokenized input ids. Can be obtained using the `__call__` method.
        skip_special_tokens (`bool`, *optional*, defaults to `False`):
            Whether or not to remove special tokens in the decoding.
        clean_up_tokenization_spaces (`bool`, *optional*):
            Whether or not to clean up the tokenization spaces. If `None`, will default to
            `self.clean_up_tokenization_spaces`.
        kwargs (additional keyword arguments, *optional*):
            Will be passed to the underlying model specific decode method.

    Returns:
        `List[str]`: The list of decoded sentences.
    """
    return [
        self.decode(
            seq,
            skip_special_tokens=skip_special_tokens,
            clean_up_tokenization_spaces=clean_up_tokenization_spaces,
            **kwargs,
        )
        for seq in sequences
    ]

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.batch_encode_plus(batch_text_or_text_pairs, add_special_tokens=True, padding=False, truncation=None, max_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, return_tensors=None, return_token_type_ids=None, return_attention_mask=None, return_overflowing_tokens=False, return_special_tokens_mask=False, return_offsets_mapping=False, return_length=False, verbose=True, split_special_tokens=False, **kwargs)

Tokenize and prepare for the model a list of sequences or a list of pairs of sequences.

This method is deprecated, __call__ should be used instead.

PARAMETER	DESCRIPTION
also	Batch of sequences or pair of sequences to be encoded. This can be a list of string/string-sequences/int-sequences or a list of pair of string/string-sequences/int-sequence (see details in encode_plus). TYPE: `List[List[int]]`, `List[Tuple[List[int], List[int]]]`

Source code in mindnlp\transformers\tokenization_utils_base.py

def batch_encode_plus(
    self,
    batch_text_or_text_pairs: Union[
        List[TextInput],
        List[TextInputPair],
        List[PreTokenizedInput],
        List[PreTokenizedInputPair],
        List[EncodedInput],
        List[EncodedInputPair],
    ],
    add_special_tokens: bool = True,
    padding: Union[bool, str, PaddingStrategy] = False,
    truncation: Union[bool, str, TruncationStrategy] = None,
    max_length: Optional[int] = None,
    stride: int = 0,
    is_split_into_words: bool = False,
    pad_to_multiple_of: Optional[int] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    return_token_type_ids: Optional[bool] = None,
    return_attention_mask: Optional[bool] = None,
    return_overflowing_tokens: bool = False,
    return_special_tokens_mask: bool = False,
    return_offsets_mapping: bool = False,
    return_length: bool = False,
    verbose: bool = True,
    split_special_tokens: bool = False,
    **kwargs,
) -> BatchEncoding:
    """
    Tokenize and prepare for the model a list of sequences or a list of pairs of sequences.

    <Tip warning={true}>

    This method is deprecated, `__call__` should be used instead.

    </Tip>

    Args:
        batch_text_or_text_pairs (`List[str]`, `List[Tuple[str, str]]`, `List[List[str]]`, `List[Tuple[List[str], List[str]]]`, and for not-fast tokenizers,
        also `List[List[int]]`, `List[Tuple[List[int], List[int]]]`):
            Batch of sequences or pair of sequences to be encoded. This can be a list of
            string/string-sequences/int-sequences or a list of pair of string/string-sequences/int-sequence (see
            details in `encode_plus`).
    """

    # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
    padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
        padding=padding,
        truncation=truncation,
        max_length=max_length,
        pad_to_multiple_of=pad_to_multiple_of,
        verbose=verbose,
        **kwargs,
    )

    return self._batch_encode_plus(
        batch_text_or_text_pairs=batch_text_or_text_pairs,
        add_special_tokens=add_special_tokens,
        padding_strategy=padding_strategy,
        truncation_strategy=truncation_strategy,
        max_length=max_length,
        stride=stride,
        is_split_into_words=is_split_into_words,
        pad_to_multiple_of=pad_to_multiple_of,
        return_tensors=return_tensors,
        return_token_type_ids=return_token_type_ids,
        return_attention_mask=return_attention_mask,
        return_overflowing_tokens=return_overflowing_tokens,
        return_special_tokens_mask=return_special_tokens_mask,
        return_offsets_mapping=return_offsets_mapping,
        return_length=return_length,
        verbose=verbose,
        split_special_tokens=split_special_tokens,
        **kwargs,
    )

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.build_inputs_with_special_tokens(token_ids_0, token_ids_1=None)

Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens.

This implementation does not add special tokens and this method should be overridden in a subclass.

PARAMETER	DESCRIPTION
token_ids_0	The first tokenized sequence. TYPE: `List[int]`
token_ids_1	The second tokenized sequence. TYPE: `List[int]`, *optional* DEFAULT: None

RETURNS	DESCRIPTION
List[int]	List[int]: The model input with special tokens.

Source code in mindnlp\transformers\tokenization_utils_base.py

def build_inputs_with_special_tokens(
    self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
) -> List[int]:
    """
    Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and
    adding special tokens.

    This implementation does not add special tokens and this method should be overridden in a subclass.

    Args:
        token_ids_0 (`List[int]`): The first tokenized sequence.
        token_ids_1 (`List[int]`, *optional*): The second tokenized sequence.

    Returns:
        `List[int]`: The model input with special tokens.
    """
    if token_ids_1 is None:
        return token_ids_0
    return token_ids_0 + token_ids_1

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.clean_up_tokenization(out_string) staticmethod

Clean up a list of simple English tokenization artifacts like spaces before punctuations and abbreviated forms.

PARAMETER	DESCRIPTION
out_string	The text to clean up. TYPE: `str`

RETURNS	DESCRIPTION
str	str: The cleaned-up string.

Source code in mindnlp\transformers\tokenization_utils_base.py

@staticmethod
def clean_up_tokenization(out_string: str) -> str:
    """
    Clean up a list of simple English tokenization artifacts like spaces before punctuations and abbreviated forms.

    Args:
        out_string (`str`): The text to clean up.

    Returns:
        `str`: The cleaned-up string.
    """
    out_string = (
        out_string.replace(" .", ".")
        .replace(" ?", "?")
        .replace(" !", "!")
        .replace(" ,", ",")
        .replace(" ' ", "'")
        .replace(" n't", "n't")
        .replace(" 'm", "'m")
        .replace(" 's", "'s")
        .replace(" 've", "'ve")
        .replace(" 're", "'re")
    )
    return out_string

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.convert_tokens_to_string(tokens)

Converts a sequence of tokens in a single string. The most simple way to do it is " ".join(tokens) but we often want to remove sub-word tokenization artifacts at the same time.

PARAMETER	DESCRIPTION
tokens	The token to join in a string. TYPE: `List[str]`

RETURNS	DESCRIPTION
str	str: The joined tokens.

Source code in mindnlp\transformers\tokenization_utils_base.py

def convert_tokens_to_string(self, tokens: List[str]) -> str:
    """
    Converts a sequence of tokens in a single string. The most simple way to do it is `" ".join(tokens)` but we
    often want to remove sub-word tokenization artifacts at the same time.

    Args:
        tokens (`List[str]`): The token to join in a string.

    Returns:
        `str`: The joined tokens.
    """
    raise NotImplementedError

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.create_token_type_ids_from_sequences(token_ids_0, token_ids_1=None)

Create the token type IDs corresponding to the sequences passed. What are token type IDs?

Should be overridden in a subclass if the model has a special way of building those.

PARAMETER	DESCRIPTION
token_ids_0	The first tokenized sequence. TYPE: `List[int]`
token_ids_1	The second tokenized sequence. TYPE: `List[int]`, *optional* DEFAULT: None

RETURNS	DESCRIPTION
List[int]	List[int]: The token type ids.

Source code in mindnlp\transformers\tokenization_utils_base.py

def create_token_type_ids_from_sequences(
    self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
) -> List[int]:
    """
    Create the token type IDs corresponding to the sequences passed. [What are token type
    IDs?](../glossary#token-type-ids)

    Should be overridden in a subclass if the model has a special way of building those.

    Args:
        token_ids_0 (`List[int]`): The first tokenized sequence.
        token_ids_1 (`List[int]`, *optional*): The second tokenized sequence.

    Returns:
        `List[int]`: The token type ids.
    """
    if token_ids_1 is None:
        return len(token_ids_0) * [0]
    return [0] * len(token_ids_0) + [1] * len(token_ids_1)

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.decode(token_ids, skip_special_tokens=False, clean_up_tokenization_spaces=None, **kwargs)

Converts a sequence of ids in a string, using the tokenizer and vocabulary with options to remove special tokens and clean up tokenization spaces.

Similar to doing self.convert_tokens_to_string(self.convert_ids_to_tokens(token_ids)).

PARAMETER	DESCRIPTION
token_ids	List of tokenized input ids. Can be obtained using the __call__ method. TYPE: `Union[int, List[int], np.ndarray, mindspore.Tensor, tf.Tensor]`
skip_special_tokens	Whether or not to remove special tokens in the decoding. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False
clean_up_tokenization_spaces	Whether or not to clean up the tokenization spaces. If None, will default to self.clean_up_tokenization_spaces. TYPE: `bool`, *optional* DEFAULT: None
kwargs	Will be passed to the underlying model specific decode method. TYPE: additional keyword arguments, *optional* DEFAULT: {}

RETURNS	DESCRIPTION
str	str: The decoded sentence.

Source code in mindnlp\transformers\tokenization_utils_base.py

def decode(
    self,
    token_ids: Union[int, List[int], "np.ndarray", "mindspore.Tensor"],
    skip_special_tokens: bool = False,
    clean_up_tokenization_spaces: bool = None,
    **kwargs,
) -> str:
    """
    Converts a sequence of ids in a string, using the tokenizer and vocabulary with options to remove special
    tokens and clean up tokenization spaces.

    Similar to doing `self.convert_tokens_to_string(self.convert_ids_to_tokens(token_ids))`.

    Args:
        token_ids (`Union[int, List[int], np.ndarray, mindspore.Tensor, tf.Tensor]`):
            List of tokenized input ids. Can be obtained using the `__call__` method.
        skip_special_tokens (`bool`, *optional*, defaults to `False`):
            Whether or not to remove special tokens in the decoding.
        clean_up_tokenization_spaces (`bool`, *optional*):
            Whether or not to clean up the tokenization spaces. If `None`, will default to
            `self.clean_up_tokenization_spaces`.
        kwargs (additional keyword arguments, *optional*):
            Will be passed to the underlying model specific decode method.

    Returns:
        `str`: The decoded sentence.
    """
    # Convert inputs to python lists
    token_ids = to_py_obj(token_ids)

    return self._decode(
        token_ids=token_ids,
        skip_special_tokens=skip_special_tokens,
        clean_up_tokenization_spaces=clean_up_tokenization_spaces,
        **kwargs,
    )

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.encode(text, text_pair=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, stride=0, return_tensors=None, **kwargs)

Converts a string to a sequence of ids (integer), using the tokenizer and vocabulary.

Same as doing self.convert_tokens_to_ids(self.tokenize(text)).

PARAMETER	DESCRIPTION
text	The first sequence to be encoded. This can be a string, a list of strings (tokenized string using the tokenize method) or a list of integers (tokenized string ids using the convert_tokens_to_ids method). TYPE: `str`, `List[str]` or `List[int]`
text_pair	Optional second sequence to be encoded. This can be a string, a list of strings (tokenized string using the tokenize method) or a list of integers (tokenized string ids using the convert_tokens_to_ids method). TYPE: `str`, `List[str]` or `List[int]`, *optional* DEFAULT: None

Source code in mindnlp\transformers\tokenization_utils_base.py

def encode(
    self,
    text: Union[TextInput, PreTokenizedInput, EncodedInput],
    text_pair: Optional[Union[TextInput, PreTokenizedInput, EncodedInput]] = None,
    add_special_tokens: bool = True,
    padding: Union[bool, str, PaddingStrategy] = False,
    truncation: Union[bool, str, TruncationStrategy] = None,
    max_length: Optional[int] = None,
    stride: int = 0,
    return_tensors: Optional[Union[str, TensorType]] = None,
    **kwargs,
) -> List[int]:
    """
    Converts a string to a sequence of ids (integer), using the tokenizer and vocabulary.

    Same as doing `self.convert_tokens_to_ids(self.tokenize(text))`.

    Args:
        text (`str`, `List[str]` or `List[int]`):
            The first sequence to be encoded. This can be a string, a list of strings (tokenized string using the
            `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
            method).
        text_pair (`str`, `List[str]` or `List[int]`, *optional*):
            Optional second sequence to be encoded. This can be a string, a list of strings (tokenized string using
            the `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
            method).
    """
    encoded_inputs = self.encode_plus(
        text,
        text_pair=text_pair,
        add_special_tokens=add_special_tokens,
        padding=padding,
        truncation=truncation,
        max_length=max_length,
        stride=stride,
        return_tensors=return_tensors,
        **kwargs,
    )

    return encoded_inputs["input_ids"]

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.encode_plus(text, text_pair=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, return_tensors=None, return_token_type_ids=None, return_attention_mask=None, return_overflowing_tokens=False, return_special_tokens_mask=False, return_offsets_mapping=False, return_length=False, verbose=True, **kwargs)

Tokenize and prepare for the model a sequence or a pair of sequences.

This method is deprecated, __call__ should be used instead.

PARAMETER	DESCRIPTION
text	The first sequence to be encoded. This can be a string, a list of strings (tokenized string using the tokenize method) or a list of integers (tokenized string ids using the convert_tokens_to_ids method). TYPE: `str`, `List[str]` or (for non-fast tokenizers) `List[int]`
text_pair	Optional second sequence to be encoded. This can be a string, a list of strings (tokenized string using the tokenize method) or a list of integers (tokenized string ids using the convert_tokens_to_ids method). TYPE: `str`, `List[str]` or `List[int]`, *optional* DEFAULT: None

Source code in mindnlp\transformers\tokenization_utils_base.py

def encode_plus(
    self,
    text: Union[TextInput, PreTokenizedInput, EncodedInput],
    text_pair: Optional[Union[TextInput, PreTokenizedInput, EncodedInput]] = None,
    add_special_tokens: bool = True,
    padding: Union[bool, str, PaddingStrategy] = False,
    truncation: Union[bool, str, TruncationStrategy] = None,
    max_length: Optional[int] = None,
    stride: int = 0,
    is_split_into_words: bool = False,
    pad_to_multiple_of: Optional[int] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    return_token_type_ids: Optional[bool] = None,
    return_attention_mask: Optional[bool] = None,
    return_overflowing_tokens: bool = False,
    return_special_tokens_mask: bool = False,
    return_offsets_mapping: bool = False,
    return_length: bool = False,
    verbose: bool = True,
    **kwargs,
) -> BatchEncoding:
    """
    Tokenize and prepare for the model a sequence or a pair of sequences.

    <Tip warning={true}>

    This method is deprecated, `__call__` should be used instead.

    </Tip>

    Args:
        text (`str`, `List[str]` or (for non-fast tokenizers) `List[int]`):
            The first sequence to be encoded. This can be a string, a list of strings (tokenized string using the
            `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
            method).
        text_pair (`str`, `List[str]` or `List[int]`, *optional*):
            Optional second sequence to be encoded. This can be a string, a list of strings (tokenized string using
            the `tokenize` method) or a list of integers (tokenized string ids using the `convert_tokens_to_ids`
            method).
    """

    # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
    padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
        padding=padding,
        truncation=truncation,
        max_length=max_length,
        pad_to_multiple_of=pad_to_multiple_of,
        verbose=verbose,
        **kwargs,
    )

    return self._encode_plus(
        text=text,
        text_pair=text_pair,
        add_special_tokens=add_special_tokens,
        padding_strategy=padding_strategy,
        truncation_strategy=truncation_strategy,
        max_length=max_length,
        stride=stride,
        is_split_into_words=is_split_into_words,
        pad_to_multiple_of=pad_to_multiple_of,
        return_tensors=return_tensors,
        return_token_type_ids=return_token_type_ids,
        return_attention_mask=return_attention_mask,
        return_overflowing_tokens=return_overflowing_tokens,
        return_special_tokens_mask=return_special_tokens_mask,
        return_offsets_mapping=return_offsets_mapping,
        return_length=return_length,
        verbose=verbose,
        split_special_tokens=kwargs.pop("split_special_tokens", self.split_special_tokens),
        **kwargs,
    )

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.from_pretrained(pretrained_model_name_or_path, *init_inputs, cache_dir=None, force_download=False, local_files_only=False, token=None, revision='main', trust_remote_code=False, **kwargs) classmethod

Instantiate a [~tokenization_utils_base.PreTrainedTokenizerBase] (or a derived class) from a predefined tokenizer.

PARAMETER	DESCRIPTION
pretrained_model_name_or_path	Can be either: A string, the model id of a predefined tokenizer hosted inside a model repo on huggingface.co. A path to a directory containing vocabulary files required by the tokenizer, for instance saved using the [~tokenization_utils_base.PreTrainedTokenizerBase.save_pretrained] method, e.g., ./my_model_directory/. (Deprecated, not applicable to all derived classes) A path or url to a single saved vocabulary file (if and only if the tokenizer only requires a single vocabulary file like Bert or XLNet), e.g., ./my_model_directory/vocab.txt. TYPE: `str` or `os.PathLike`
cache_dir	Path to a directory in which a downloaded predefined tokenizer vocabulary files should be cached if the standard cache should not be used. TYPE: `str` or `os.PathLike`, *optional* DEFAULT: None
force_download	Whether or not to force the (re-)download the vocabulary files and override the cached versions if they exist. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False
resume_download	Deprecated and ignored. All downloads are now resumed by default when possible. Will be removed in v5 of Transformers.
proxies	A dictionary of proxy servers to use by protocol or endpoint, e.g., {'http': 'foo.bar:3128', 'http://hostname': 'foo.bar:4012'}. The proxies are used on each request. TYPE: `Dict[str, str]`, *optional*
token	The token to use as HTTP bearer authorization for remote files. If True, will use the token generated when running huggingface-cli login (stored in ~/.huggingface). TYPE: `str` or *bool*, *optional* DEFAULT: None
local_files_only	Whether or not to only rely on local files and not to attempt to download any files. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False
revision	The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a git-based system for storing models and other artifacts on huggingface.co, so revision can be any identifier allowed by git. TYPE: `str`, *optional*, defaults to `"main"` DEFAULT: 'main'
subfolder	In case the relevant files are located inside a subfolder of the model repo on huggingface.co (e.g. for facebook/rag-token-base), specify it here. TYPE: `str`, *optional*
inputs	Will be passed along to the Tokenizer __init__ method. TYPE: additional positional arguments, *optional*
trust_remote_code	Whether or not to allow for custom models defined on the Hub in their own modeling files. This option should only be set to True for repositories you trust and in which you have read the code, as it will execute code present on the Hub on your local machine. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False
kwargs	Will be passed to the Tokenizer __init__ method. Can be used to set special tokens like bos_token, eos_token, unk_token, sep_token, pad_token, cls_token, mask_token, additional_special_tokens. See parameters in the __init__ for more details. TYPE: additional keyword arguments, *optional* DEFAULT: {}

Passing token=True is required when you want to use a private model.

Examples:

# We can't instantiate directly the base class *PreTrainedTokenizerBase* so let's show our examples on a derived class: BertTokenizer
# Download vocabulary from huggingface.co and cache.
tokenizer = BertTokenizer.from_pretrained("google-bert/bert-base-uncased")

# Download vocabulary from huggingface.co (user-uploaded) and cache.
tokenizer = BertTokenizer.from_pretrained("dbmdz/bert-base-german-cased")

# If vocabulary files are in a directory (e.g. tokenizer was saved using *save_pretrained('./test/saved_model/')*)
tokenizer = BertTokenizer.from_pretrained("./test/saved_model/")

# If the tokenizer uses a single vocabulary file, you can point directly to this file
tokenizer = BertTokenizer.from_pretrained("./test/saved_model/my_vocab.txt")

# You can link tokens to special vocabulary when instantiating
tokenizer = BertTokenizer.from_pretrained("google-bert/bert-base-uncased", unk_token="<unk>")
# You should be sure '<unk>' is in the vocabulary when doing that.
# Otherwise use tokenizer.add_special_tokens({'unk_token': '<unk>'}) instead)
assert tokenizer.unk_token == "<unk>"

Source code in mindnlp\transformers\tokenization_utils_base.py

@classmethod
def from_pretrained(
    cls,
    pretrained_model_name_or_path: Union[str, os.PathLike],
    *init_inputs,
    cache_dir: Optional[Union[str, os.PathLike]] = None,
    force_download: bool = False,
    local_files_only: bool = False,
    token: Optional[Union[str, bool]] = None,
    revision: str = "main",
    trust_remote_code=False,
    **kwargs,
):
    r"""
    Instantiate a [`~tokenization_utils_base.PreTrainedTokenizerBase`] (or a derived class) from a predefined
    tokenizer.

    Args:
        pretrained_model_name_or_path (`str` or `os.PathLike`):
            Can be either:

            - A string, the *model id* of a predefined tokenizer hosted inside a model repo on huggingface.co.
            - A path to a *directory* containing vocabulary files required by the tokenizer, for instance saved
              using the [`~tokenization_utils_base.PreTrainedTokenizerBase.save_pretrained`] method, e.g.,
              `./my_model_directory/`.
            - (**Deprecated**, not applicable to all derived classes) A path or url to a single saved vocabulary
              file (if and only if the tokenizer only requires a single vocabulary file like Bert or XLNet), e.g.,
              `./my_model_directory/vocab.txt`.
        cache_dir (`str` or `os.PathLike`, *optional*):
            Path to a directory in which a downloaded predefined tokenizer vocabulary files should be cached if the
            standard cache should not be used.
        force_download (`bool`, *optional*, defaults to `False`):
            Whether or not to force the (re-)download the vocabulary files and override the cached versions if they
            exist.
        resume_download:
            Deprecated and ignored. All downloads are now resumed by default when possible.
            Will be removed in v5 of Transformers.
        proxies (`Dict[str, str]`, *optional*):
            A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
            'http://hostname': 'foo.bar:4012'}`. The proxies are used on each request.
        token (`str` or *bool*, *optional*):
            The token to use as HTTP bearer authorization for remote files. If `True`, will use the token generated
            when running `huggingface-cli login` (stored in `~/.huggingface`).
        local_files_only (`bool`, *optional*, defaults to `False`):
            Whether or not to only rely on local files and not to attempt to download any files.
        revision (`str`, *optional*, defaults to `"main"`):
            The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a
            git-based system for storing models and other artifacts on huggingface.co, so `revision` can be any
            identifier allowed by git.
        subfolder (`str`, *optional*):
            In case the relevant files are located inside a subfolder of the model repo on huggingface.co (e.g. for
            facebook/rag-token-base), specify it here.
        inputs (additional positional arguments, *optional*):
            Will be passed along to the Tokenizer `__init__` method.
        trust_remote_code (`bool`, *optional*, defaults to `False`):
            Whether or not to allow for custom models defined on the Hub in their own modeling files. This option
            should only be set to `True` for repositories you trust and in which you have read the code, as it will
            execute code present on the Hub on your local machine.
        kwargs (additional keyword arguments, *optional*):
            Will be passed to the Tokenizer `__init__` method. Can be used to set special tokens like `bos_token`,
            `eos_token`, `unk_token`, `sep_token`, `pad_token`, `cls_token`, `mask_token`,
            `additional_special_tokens`. See parameters in the `__init__` for more details.

    <Tip>

    Passing `token=True` is required when you want to use a private model.

    </Tip>

    Examples:

    ```python
    # We can't instantiate directly the base class *PreTrainedTokenizerBase* so let's show our examples on a derived class: BertTokenizer
    # Download vocabulary from huggingface.co and cache.
    tokenizer = BertTokenizer.from_pretrained("google-bert/bert-base-uncased")

    # Download vocabulary from huggingface.co (user-uploaded) and cache.
    tokenizer = BertTokenizer.from_pretrained("dbmdz/bert-base-german-cased")

    # If vocabulary files are in a directory (e.g. tokenizer was saved using *save_pretrained('./test/saved_model/')*)
    tokenizer = BertTokenizer.from_pretrained("./test/saved_model/")

    # If the tokenizer uses a single vocabulary file, you can point directly to this file
    tokenizer = BertTokenizer.from_pretrained("./test/saved_model/my_vocab.txt")

    # You can link tokens to special vocabulary when instantiating
    tokenizer = BertTokenizer.from_pretrained("google-bert/bert-base-uncased", unk_token="<unk>")
    # You should be sure '<unk>' is in the vocabulary when doing that.
    # Otherwise use tokenizer.add_special_tokens({'unk_token': '<unk>'}) instead)
    assert tokenizer.unk_token == "<unk>"
    ```"""
    resume_download = kwargs.pop("resume_download", None)
    proxies = kwargs.pop("proxies", None)
    use_auth_token = kwargs.pop("use_auth_token", None)
    subfolder = kwargs.pop("subfolder", None)
    from_pipeline = kwargs.pop("_from_pipeline", None)
    from_auto_class = kwargs.pop("_from_auto", False)
    commit_hash = kwargs.pop("_commit_hash", None)
    gguf_file = kwargs.get("gguf_file", None)
    mirror = kwargs.get("mirror", "huggingface")

    if use_auth_token is not None:
        warnings.warn(
            "The `use_auth_token` argument is deprecated. Please use `token` instead.",
            FutureWarning,
        )
        if token is not None:
            raise ValueError(
                "`token` and `use_auth_token` are both specified. Please set only the argument `token`."
            )
        token = use_auth_token

    user_agent = {"file_type": "tokenizer", "from_auto_class": from_auto_class, "is_fast": "Fast" in cls.__name__}
    if from_pipeline is not None:
        user_agent["using_pipeline"] = from_pipeline

    if is_offline_mode() and not local_files_only:
        logger.info("Offline mode: forcing local_files_only=True")
        local_files_only = True

    pretrained_model_name_or_path = str(pretrained_model_name_or_path)
    vocab_files = {}
    init_configuration = {}

    is_local = os.path.isdir(pretrained_model_name_or_path)
    single_file_id = None
    if os.path.isfile(pretrained_model_name_or_path) or is_remote_url(pretrained_model_name_or_path):
        if len(cls.vocab_files_names) > 1 and not gguf_file:
            raise ValueError(
                f"Calling {cls.__name__}.from_pretrained() with the path to a single file or url is not "
                "supported for this tokenizer. Use a model identifier or the path to a directory instead."
            )
        warnings.warn(
            f"Calling {cls.__name__}.from_pretrained() with the path to a single file or url is deprecated and "
            "won't be possible anymore in v5. Use a model identifier or the path to a directory instead.",
            FutureWarning,
        )
        file_id = list(cls.vocab_files_names.keys())[0]

        vocab_files[file_id] = pretrained_model_name_or_path
        single_file_id = file_id
    else:
        if gguf_file:
            vocab_files["vocab_file"] = gguf_file
        else:
            # At this point pretrained_model_name_or_path is either a directory or a model identifier name
            additional_files_names = {
                "added_tokens_file": ADDED_TOKENS_FILE,  # kept only for legacy
                "special_tokens_map_file": SPECIAL_TOKENS_MAP_FILE,  # kept only for legacy
                "tokenizer_config_file": TOKENIZER_CONFIG_FILE,
                # tokenizer_file used to initialize a slow from a fast. Properly copy the `addedTokens` instead of adding in random orders
                "tokenizer_file": FULL_TOKENIZER_FILE,
            }
            vocab_files = {**cls.vocab_files_names, **additional_files_names}
            if "tokenizer_file" in vocab_files:
                # Try to get the tokenizer config to see if there are versioned tokenizer files.
                fast_tokenizer_file = FULL_TOKENIZER_FILE
                resolved_config_file = cached_file(
                    pretrained_model_name_or_path,
                    TOKENIZER_CONFIG_FILE,
                    cache_dir=cache_dir,
                    force_download=force_download,
                    resume_download=resume_download,
                    proxies=proxies,
                    token=token,
                    revision=revision,
                    local_files_only=local_files_only,
                    subfolder=subfolder,
                    mirror=mirror,
                    user_agent=user_agent,
                    _raise_exceptions_for_gated_repo=False,
                    _raise_exceptions_for_missing_entries=False,
                    _raise_exceptions_for_connection_errors=False,
                )
                commit_hash = extract_commit_hash(resolved_config_file, commit_hash)
                if resolved_config_file is not None:
                    with open(resolved_config_file, encoding="utf-8") as reader:
                        tokenizer_config = json.load(reader)
                        if "fast_tokenizer_files" in tokenizer_config:
                            fast_tokenizer_file = get_fast_tokenizer_file(tokenizer_config["fast_tokenizer_files"])
                vocab_files["tokenizer_file"] = fast_tokenizer_file

    # Get files from url, cache, or disk depending on the case
    resolved_vocab_files = {}
    unresolved_files = []
    for file_id, file_path in vocab_files.items():
        if file_path is None:
            resolved_vocab_files[file_id] = None
        elif single_file_id == file_id:
            if os.path.isfile(file_path):
                resolved_vocab_files[file_id] = file_path
            elif is_remote_url(file_path):
                resolved_vocab_files[file_id] = download_url(file_path, proxies=proxies)
        else:
            resolved_vocab_files[file_id] = cached_file(
                pretrained_model_name_or_path,
                file_path,
                cache_dir=cache_dir,
                force_download=force_download,
                proxies=proxies,
                resume_download=resume_download,
                local_files_only=local_files_only,
                token=token,
                user_agent=user_agent,
                revision=revision,
                subfolder=subfolder,
                mirror=mirror,
                _raise_exceptions_for_gated_repo=False,
                _raise_exceptions_for_missing_entries=False,
                _raise_exceptions_for_connection_errors=False,
                # _commit_hash=commit_hash,
            )
            commit_hash = extract_commit_hash(resolved_vocab_files[file_id], commit_hash)

    if len(unresolved_files) > 0:
        logger.info(
            f"Can't load following files from cache: {unresolved_files} and cannot check if these "
            "files are necessary for the tokenizer to operate."
        )

    # If one passes a GGUF file path to `gguf_file` there is no need for this check as the tokenizer will be
    # loaded directly from the GGUF file.
    if all(full_file_name is None for full_file_name in resolved_vocab_files.values()) and not gguf_file:
        raise EnvironmentError(
            f"Can't load tokenizer for '{pretrained_model_name_or_path}'. If you were trying to load it from "
            "'https://huggingface.co/models', make sure you don't have a local directory with the same name. "
            f"Otherwise, make sure '{pretrained_model_name_or_path}' is the correct path to a directory "
            f"containing all relevant files for a {cls.__name__} tokenizer."
        )

    for file_id, file_path in vocab_files.items():
        if file_id not in resolved_vocab_files:
            continue

        if is_local:
            logger.info(f"loading file {file_path}")
        else:
            logger.info(f"loading file {file_path} from cache at {resolved_vocab_files[file_id]}")

    return cls._from_pretrained(
        resolved_vocab_files,
        pretrained_model_name_or_path,
        init_configuration,
        *init_inputs,
        token=token,
        cache_dir=cache_dir,
        local_files_only=local_files_only,
        _commit_hash=commit_hash,
        _is_local=is_local,
        trust_remote_code=trust_remote_code,
        **kwargs,
    )

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.get_chat_template(chat_template=None, tools=None)

Retrieve the chat template string used for tokenizing chat messages. This template is used internally by the apply_chat_template method and can also be used externally to retrieve the model's chat template for better generation tracking.

PARAMETER	DESCRIPTION
chat_template	A Jinja template or the name of a template to use for this conversion. It is usually not necessary to pass anything to this argument, as the model's template will be used by default. TYPE: `str`, *optional* DEFAULT: None
tools	A list of tools (callable functions) that will be accessible to the model. If the template does not support function calling, this argument will have no effect. Each tool should be passed as a JSON Schema, giving the name, description and argument types for the tool. See our chat templating guide for more information. TYPE: `List[Dict]`, *optional* DEFAULT: None

RETURNS	DESCRIPTION
str	str: The chat template string.

Source code in mindnlp\transformers\tokenization_utils_base.py

def get_chat_template(self, chat_template: Optional[str] = None, tools: Optional[List[Dict]] = None) -> str:
    """
    Retrieve the chat template string used for tokenizing chat messages. This template is used
    internally by the `apply_chat_template` method and can also be used externally to retrieve the model's chat
    template for better generation tracking.

    Args:
        chat_template (`str`, *optional*):
            A Jinja template or the name of a template to use for this conversion.
            It is usually not necessary to pass anything to this argument,
            as the model's template will be used by default.
        tools (`List[Dict]`, *optional*):
            A list of tools (callable functions) that will be accessible to the model. If the template does not
            support function calling, this argument will have no effect. Each tool should be passed as a JSON Schema,
            giving the name, description and argument types for the tool. See our
            [chat templating guide](https://huggingface.co/docs/transformers/main/en/chat_templating#automated-function-conversion-for-tool-use)
            for more information.

    Returns:
        `str`: The chat template string.
    """
    # First, handle the cases when the model has a dict of multiple templates
    if isinstance(self.chat_template, dict):
        template_dict = self.chat_template
        if chat_template is not None and chat_template in template_dict:
            # The user can pass the name of a template to the chat template argument instead of an entire template
            chat_template = template_dict[chat_template]
        elif chat_template is None:
            if tools is not None and "tool_use" in template_dict:
                chat_template = template_dict["tool_use"]
            elif "default" in template_dict:
                chat_template = template_dict["default"]
            else:
                raise ValueError(
                    "This model has multiple chat templates with no default specified! Please either pass a chat "
                    "template or the name of the template you wish to use to the `chat_template` argument. Available "
                    f"template names are {sorted(template_dict.keys())}."
                )

    elif chat_template is None:
        # These are the cases when the model has a single template
        # priority: `chat_template` argument > `tokenizer.chat_template`
        if self.chat_template is not None:
            chat_template = self.chat_template

        else:
            raise ValueError(
                "Cannot use apply_chat_template() because tokenizer.chat_template is not set and no template "
                "argument was passed! For information about writing templates and setting the "
                "tokenizer.chat_template attribute, please see the documentation at "
                "https://huggingface.co/docs/transformers/main/en/chat_templating"
            )

    return chat_template

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.get_special_tokens_mask(token_ids_0, token_ids_1=None, already_has_special_tokens=False)

Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding special tokens using the tokenizer prepare_for_model or encode_plus methods.

PARAMETER	DESCRIPTION
token_ids_0	List of ids of the first sequence. TYPE: `List[int]`
token_ids_1	List of ids of the second sequence. TYPE: `List[int]`, *optional* DEFAULT: None
already_has_special_tokens	Whether or not the token list is already formatted with special tokens for the model. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False

RETURNS	DESCRIPTION
List[int]	A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.

Source code in mindnlp\transformers\tokenization_utils_base.py

def get_special_tokens_mask(
    self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
) -> List[int]:
    """
    Retrieves sequence ids from a token list that has no special tokens added. This method is called when adding
    special tokens using the tokenizer `prepare_for_model` or `encode_plus` methods.

    Args:
        token_ids_0 (`List[int]`):
            List of ids of the first sequence.
        token_ids_1 (`List[int]`, *optional*):
            List of ids of the second sequence.
        already_has_special_tokens (`bool`, *optional*, defaults to `False`):
            Whether or not the token list is already formatted with special tokens for the model.

    Returns:
        A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
    """
    assert already_has_special_tokens and token_ids_1 is None, (
        "You cannot use ``already_has_special_tokens=False`` with this tokenizer. "
        "Please use a slow (full python) tokenizer to activate this argument. "
        "Or set `return_special_tokens_mask=True` when calling the encoding method "
        "to get the special tokens mask in any tokenizer. "
    )

    all_special_ids = self.all_special_ids  # cache the property

    special_tokens_mask = [1 if token in all_special_ids else 0 for token in token_ids_0]

    return special_tokens_mask

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.get_vocab()

Returns the vocabulary as a dictionary of token to index.

tokenizer.get_vocab()[token] is equivalent to tokenizer.convert_tokens_to_ids(token) when token is in the vocab.

RETURNS	DESCRIPTION
Dict[str, int]	Dict[str, int]: The vocabulary.

Source code in mindnlp\transformers\tokenization_utils_base.py

def get_vocab(self) -> Dict[str, int]:
    """
    Returns the vocabulary as a dictionary of token to index.

    `tokenizer.get_vocab()[token]` is equivalent to `tokenizer.convert_tokens_to_ids(token)` when `token` is in the
    vocab.

    Returns:
        `Dict[str, int]`: The vocabulary.
    """
    raise NotImplementedError()

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.pad(encoded_inputs, padding=True, max_length=None, pad_to_multiple_of=None, return_attention_mask=None, return_tensors=None, verbose=True)

Pad a single encoded input or a batch of encoded inputs up to predefined length or to the max sequence length in the batch.

Padding side (left/right) padding token ids are defined at the tokenizer level (with self.padding_side, self.pad_token_id and self.pad_token_type_id).

Please note that with a fast tokenizer, using the __call__ method is faster than using a method to encode the text followed by a call to the pad method to get a padded encoding.

If the encoded_inputs passed are dictionary of numpy arrays, PyTorch tensors or TensorFlow tensors, the result will use the same type unless you provide a different tensor type with return_tensors. In the case of PyTorch tensors, you will lose the specific device of your tensors however.

PARAMETER	DESCRIPTION
encoded_inputs	Tokenized inputs. Can represent one input ([BatchEncoding] or Dict[str, List[int]]) or a batch of tokenized inputs (list of [BatchEncoding], Dict[str, List[List[int]]] or List[Dict[str, List[int]]]) so you can use this method during preprocessing as well as in a PyTorch Dataloader collate function. Instead of List[int] you can have tensors (numpy arrays, PyTorch tensors or TensorFlow tensors), see the note above for the return type. TYPE: [`BatchEncoding`], list of [`BatchEncoding`], `Dict[str, List[int]]`, `Dict[str, List[List[int]]` or `List[Dict[str, List[int]]]`
padding	Select a strategy to pad the returned sequences (according to the model's padding side and padding index) among: True or 'longest': Pad to the longest sequence in the batch (or no padding if only a single sequence if provided). 'max_length': Pad to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. False or 'do_not_pad' (default): No padding (i.e., can output a batch with sequences of different lengths). TYPE: `bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `True` DEFAULT: True
max_length	Maximum length of the returned list and optionally padding length (see above). TYPE: `int`, *optional* DEFAULT: None
pad_to_multiple_of	If set will pad the sequence to a multiple of the provided value. This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability >= 7.5 (Volta). TYPE: `int`, *optional* DEFAULT: None
return_attention_mask	Whether to return the attention mask. If left to the default, will return the attention mask according to the specific tokenizer's default, defined by the return_outputs attribute. What are attention masks? TYPE: `bool`, *optional* DEFAULT: None
return_tensors	If set, will return tensors instead of list of python integers. Acceptable values are: 'tf': Return TensorFlow tf.constant objects. 'pt': Return PyTorch mindspore.Tensor objects. 'np': Return Numpy np.ndarray objects. TYPE: `str` or [`~utils.TensorType`], *optional* DEFAULT: None
verbose	Whether or not to print more information and warnings. TYPE: `bool`, *optional*, defaults to `True` DEFAULT: True

Source code in mindnlp\transformers\tokenization_utils_base.py

def pad(
    self,
    encoded_inputs: Union[
        BatchEncoding,
        List[BatchEncoding],
        Dict[str, EncodedInput],
        Dict[str, List[EncodedInput]],
        List[Dict[str, EncodedInput]],
    ],
    padding: Union[bool, str, PaddingStrategy] = True,
    max_length: Optional[int] = None,
    pad_to_multiple_of: Optional[int] = None,
    return_attention_mask: Optional[bool] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    verbose: bool = True,
) -> BatchEncoding:
    """
    Pad a single encoded input or a batch of encoded inputs up to predefined length or to the max sequence length
    in the batch.

    Padding side (left/right) padding token ids are defined at the tokenizer level (with `self.padding_side`,
    `self.pad_token_id` and `self.pad_token_type_id`).

    Please note that with a fast tokenizer, using the `__call__` method is faster than using a method to encode the
    text followed by a call to the `pad` method to get a padded encoding.

    <Tip>

    If the `encoded_inputs` passed are dictionary of numpy arrays, PyTorch tensors or TensorFlow tensors, the
    result will use the same type unless you provide a different tensor type with `return_tensors`. In the case of
    PyTorch tensors, you will lose the specific device of your tensors however.

    </Tip>

    Args:
        encoded_inputs ([`BatchEncoding`], list of [`BatchEncoding`], `Dict[str, List[int]]`, `Dict[str, List[List[int]]` or `List[Dict[str, List[int]]]`):
            Tokenized inputs. Can represent one input ([`BatchEncoding`] or `Dict[str, List[int]]`) or a batch of
            tokenized inputs (list of [`BatchEncoding`], *Dict[str, List[List[int]]]* or *List[Dict[str,
            List[int]]]*) so you can use this method during preprocessing as well as in a PyTorch Dataloader
            collate function.

            Instead of `List[int]` you can have tensors (numpy arrays, PyTorch tensors or TensorFlow tensors), see
            the note above for the return type.
        padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `True`):
             Select a strategy to pad the returned sequences (according to the model's padding side and padding
             index) among:

            - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
              sequence if provided).
            - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum
              acceptable input length for the model if that argument is not provided.
            - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different
              lengths).
        max_length (`int`, *optional*):
            Maximum length of the returned list and optionally padding length (see above).
        pad_to_multiple_of (`int`, *optional*):
            If set will pad the sequence to a multiple of the provided value.

            This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability
            `>= 7.5` (Volta).
        return_attention_mask (`bool`, *optional*):
            Whether to return the attention mask. If left to the default, will return the attention mask according
            to the specific tokenizer's default, defined by the `return_outputs` attribute.

            [What are attention masks?](../glossary#attention-mask)
        return_tensors (`str` or [`~utils.TensorType`], *optional*):
            If set, will return tensors instead of list of python integers. Acceptable values are:

            - `'tf'`: Return TensorFlow `tf.constant` objects.
            - `'pt'`: Return PyTorch `mindspore.Tensor` objects.
            - `'np'`: Return Numpy `np.ndarray` objects.
        verbose (`bool`, *optional*, defaults to `True`):
            Whether or not to print more information and warnings.
    """
    if self.__class__.__name__.endswith("Fast"):
        if not self.deprecation_warnings.get("Asking-to-pad-a-fast-tokenizer", False):
            logger.warning_advice(
                f"You're using a {self.__class__.__name__} tokenizer. Please note that with a fast tokenizer,"
                " using the `__call__` method is faster than using a method to encode the text followed by a call"
                " to the `pad` method to get a padded encoding."
            )
            self.deprecation_warnings["Asking-to-pad-a-fast-tokenizer"] = True

    # If we have a list of dicts, let's convert it in a dict of lists
    # We do this to allow using this method as a collate_fn function in PyTorch Dataloader
    if isinstance(encoded_inputs, (list, tuple)) and isinstance(encoded_inputs[0], Mapping):
        encoded_inputs = {key: [example[key] for example in encoded_inputs] for key in encoded_inputs[0].keys()}

    # The model's main input name, usually `input_ids`, has be passed for padding
    if self.model_input_names[0] not in encoded_inputs:
        raise ValueError(
            "You should supply an encoding or a list of encodings to this method "
            f"that includes {self.model_input_names[0]}, but you provided {list(encoded_inputs.keys())}"
        )

    required_input = encoded_inputs[self.model_input_names[0]]

    if required_input is None or (isinstance(required_input, Sized) and len(required_input) == 0):
        if return_attention_mask:
            encoded_inputs["attention_mask"] = []
        return encoded_inputs

    # If we have PyTorch/TF/NumPy tensors/arrays as inputs, we cast them as python objects
    # and rebuild them afterwards if no return_tensors is specified
    # Note that we lose the specific device the tensor may be on for PyTorch

    first_element = required_input[0]
    if isinstance(first_element, (list, tuple)):
        # first_element might be an empty list/tuple in some edge cases so we grab the first non empty element.
        for item in required_input:
            if len(item) != 0:
                first_element = item[0]
                break
    # At this state, if `first_element` is still a list/tuple, it's an empty one so there is nothing to do.
    if not isinstance(first_element, (int, list, tuple)):
        if isinstance(first_element, mindspore.Tensor):
            return_tensors = "ms" if return_tensors is None else return_tensors
        elif isinstance(first_element, np.ndarray):
            return_tensors = "np" if return_tensors is None else return_tensors
        else:
            raise ValueError(
                f"type of {first_element} unknown: {type(first_element)}. "
                "Should be one of a python, numpy, mindsporeobject."
            )

        for key, value in encoded_inputs.items():
            encoded_inputs[key] = to_py_obj(value)

    # Convert padding_strategy in PaddingStrategy
    padding_strategy, _, max_length, _ = self._get_padding_truncation_strategies(
        padding=padding, max_length=max_length, verbose=verbose
    )

    required_input = encoded_inputs[self.model_input_names[0]]
    if required_input and not isinstance(required_input[0], (list, tuple)):
        encoded_inputs = self._pad(
            encoded_inputs,
            max_length=max_length,
            padding_strategy=padding_strategy,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
        )
        return BatchEncoding(encoded_inputs, tensor_type=return_tensors)

    batch_size = len(required_input)
    assert all(
        len(v) == batch_size for v in encoded_inputs.values()
    ), "Some items in the output dictionary have a different batch size than others."

    if padding_strategy == PaddingStrategy.LONGEST:
        max_length = max(len(inputs) for inputs in required_input)
        padding_strategy = PaddingStrategy.MAX_LENGTH

    batch_outputs = {}
    for i in range(batch_size):
        inputs = {k: v[i] for k, v in encoded_inputs.items()}
        outputs = self._pad(
            inputs,
            max_length=max_length,
            padding_strategy=padding_strategy,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
        )

        for key, value in outputs.items():
            if key not in batch_outputs:
                batch_outputs[key] = []
            batch_outputs[key].append(value)

    return BatchEncoding(batch_outputs, tensor_type=return_tensors)

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.prepare_for_model(ids, pair_ids=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, stride=0, pad_to_multiple_of=None, return_tensors=None, return_token_type_ids=None, return_attention_mask=None, return_overflowing_tokens=False, return_special_tokens_mask=False, return_offsets_mapping=False, return_length=False, verbose=True, prepend_batch_axis=False, **kwargs)

Prepares a sequence of input id, or a pair of sequences of inputs ids so that it can be used by the model. It adds special tokens, truncates sequences if overflowing while taking into account the special tokens and manages a moving window (with user defined stride) for overflowing tokens. Please Note, for pair_ids different than None and truncation_strategy = longest_first or True, it is not possible to return overflowing tokens. Such a combination of arguments will raise an error.

PARAMETER	DESCRIPTION
ids	Tokenized input ids of the first sequence. Can be obtained from a string by chaining the tokenize and convert_tokens_to_ids methods. TYPE: `List[int]`
pair_ids	Tokenized input ids of the second sequence. Can be obtained from a string by chaining the tokenize and convert_tokens_to_ids methods. TYPE: `List[int]`, *optional* DEFAULT: None

Source code in mindnlp\transformers\tokenization_utils_base.py

def prepare_for_model(
    self,
    ids: List[int],
    pair_ids: Optional[List[int]] = None,
    add_special_tokens: bool = True,
    padding: Union[bool, str, PaddingStrategy] = False,
    truncation: Union[bool, str, TruncationStrategy] = None,
    max_length: Optional[int] = None,
    stride: int = 0,
    pad_to_multiple_of: Optional[int] = None,
    return_tensors: Optional[Union[str, TensorType]] = None,
    return_token_type_ids: Optional[bool] = None,
    return_attention_mask: Optional[bool] = None,
    return_overflowing_tokens: bool = False,
    return_special_tokens_mask: bool = False,
    return_offsets_mapping: bool = False,
    return_length: bool = False,
    verbose: bool = True,
    prepend_batch_axis: bool = False,
    **kwargs,
) -> BatchEncoding:
    """
    Prepares a sequence of input id, or a pair of sequences of inputs ids so that it can be used by the model. It
    adds special tokens, truncates sequences if overflowing while taking into account the special tokens and
    manages a moving window (with user defined stride) for overflowing tokens. Please Note, for *pair_ids*
    different than `None` and *truncation_strategy = longest_first* or `True`, it is not possible to return
    overflowing tokens. Such a combination of arguments will raise an error.

    Args:
        ids (`List[int]`):
            Tokenized input ids of the first sequence. Can be obtained from a string by chaining the `tokenize` and
            `convert_tokens_to_ids` methods.
        pair_ids (`List[int]`, *optional*):
            Tokenized input ids of the second sequence. Can be obtained from a string by chaining the `tokenize`
            and `convert_tokens_to_ids` methods.
    """

    # Backward compatibility for 'truncation_strategy', 'pad_to_max_length'
    padding_strategy, truncation_strategy, max_length, kwargs = self._get_padding_truncation_strategies(
        padding=padding,
        truncation=truncation,
        max_length=max_length,
        pad_to_multiple_of=pad_to_multiple_of,
        verbose=verbose,
        **kwargs,
    )

    pair = bool(pair_ids is not None)
    len_ids = len(ids)
    len_pair_ids = len(pair_ids) if pair else 0

    if return_token_type_ids and not add_special_tokens:
        raise ValueError(
            "Asking to return token_type_ids while setting add_special_tokens to False "
            "results in an undefined behavior. Please set add_special_tokens to True or "
            "set return_token_type_ids to None."
        )

    if (
        return_overflowing_tokens
        and truncation_strategy == TruncationStrategy.LONGEST_FIRST
        and pair_ids is not None
    ):
        raise ValueError(
            "Not possible to return overflowing tokens for pair of sequences with the "
            "`longest_first`. Please select another truncation strategy than `longest_first`, "
            "for instance `only_second` or `only_first`."
        )

    # Load from model defaults
    if return_token_type_ids is None:
        return_token_type_ids = "token_type_ids" in self.model_input_names
    if return_attention_mask is None:
        return_attention_mask = "attention_mask" in self.model_input_names

    encoded_inputs = {}

    # Compute the total size of the returned encodings
    total_len = len_ids + len_pair_ids + (self.num_special_tokens_to_add(pair=pair) if add_special_tokens else 0)

    # Truncation: Handle max sequence length
    overflowing_tokens = []
    if truncation_strategy != TruncationStrategy.DO_NOT_TRUNCATE and max_length and total_len > max_length:
        ids, pair_ids, overflowing_tokens = self.truncate_sequences(
            ids,
            pair_ids=pair_ids,
            num_tokens_to_remove=total_len - max_length,
            truncation_strategy=truncation_strategy,
            stride=stride,
        )

    if return_overflowing_tokens:
        encoded_inputs["overflowing_tokens"] = overflowing_tokens
        encoded_inputs["num_truncated_tokens"] = total_len - max_length

    # Add special tokens
    if add_special_tokens:
        sequence = self.build_inputs_with_special_tokens(ids, pair_ids)
        token_type_ids = self.create_token_type_ids_from_sequences(ids, pair_ids)
    else:
        sequence = ids + pair_ids if pair else ids
        token_type_ids = [0] * len(ids) + ([0] * len(pair_ids) if pair else [])

    # Build output dictionary
    encoded_inputs["input_ids"] = sequence
    if return_token_type_ids:
        encoded_inputs["token_type_ids"] = token_type_ids
    if return_special_tokens_mask:
        if add_special_tokens:
            encoded_inputs["special_tokens_mask"] = self.get_special_tokens_mask(ids, pair_ids)
        else:
            encoded_inputs["special_tokens_mask"] = [0] * len(sequence)

    # Check lengths
    self._eventual_warn_about_too_long_sequence(encoded_inputs["input_ids"], max_length, verbose)

    # Padding
    if padding_strategy != PaddingStrategy.DO_NOT_PAD or return_attention_mask:
        encoded_inputs = self.pad(
            encoded_inputs,
            max_length=max_length,
            padding=padding_strategy.value,
            pad_to_multiple_of=pad_to_multiple_of,
            return_attention_mask=return_attention_mask,
        )

    if return_length:
        encoded_inputs["length"] = len(encoded_inputs["input_ids"])

    batch_outputs = BatchEncoding(
        encoded_inputs, tensor_type=return_tensors, prepend_batch_axis=prepend_batch_axis
    )

    return batch_outputs

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.prepare_seq2seq_batch(src_texts, tgt_texts=None, max_length=None, max_target_length=None, padding='longest', return_tensors=None, truncation=True, **kwargs)

Prepare model inputs for translation. For best performance, translate one sentence at a time.

PARAMETER	DESCRIPTION
src_texts	List of documents to summarize or source language texts. TYPE: `List[str]`
tgt_texts	List of summaries or target language texts. TYPE: `list`, *optional* DEFAULT: None
max_length	Controls the maximum length for encoder inputs (documents to summarize or source language texts) If left unset or set to None, this will use the predefined model maximum length if a maximum length is required by one of the truncation/padding parameters. If the model has no specific maximum input length (like XLNet) truncation/padding to a maximum length will be deactivated. TYPE: `int`, *optional* DEFAULT: None
max_target_length	Controls the maximum length of decoder inputs (target language texts or summaries) If left unset or set to None, this will use the max_length value. TYPE: `int`, *optional* DEFAULT: None
padding	Activates and controls padding. Accepts the following values: True or 'longest': Pad to the longest sequence in the batch (or no padding if only a single sequence if provided). 'max_length': Pad to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. False or 'do_not_pad' (default): No padding (i.e., can output a batch with sequences of different lengths). TYPE: `bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False` DEFAULT: 'longest'
return_tensors	If set, will return tensors instead of list of python integers. Acceptable values are: 'tf': Return TensorFlow tf.constant objects. 'pt': Return PyTorch mindspore.Tensor objects. 'np': Return Numpy np.ndarray objects. TYPE: `str` or [`~utils.TensorType`], *optional* DEFAULT: None
truncation	Activates and controls truncation. Accepts the following values: True or 'longest_first': Truncate to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. This will truncate token by token, removing a token from the longest sequence in the pair if a pair of sequences (or a batch of pairs) is provided. 'only_first': Truncate to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. This will only truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided. 'only_second': Truncate to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. This will only truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided. False or 'do_not_truncate' (default): No truncation (i.e., can output batch with sequence lengths greater than the model maximum admissible input size). TYPE: `bool`, `str` or [`~tokenization_utils_base.TruncationStrategy`], *optional*, defaults to `True` DEFAULT: True
**kwargs	Additional keyword arguments passed along to self.__call__. DEFAULT: {}

Return

[BatchEncoding]: A [BatchEncoding] with the following fields:

• input_ids -- List of token ids to be fed to the encoder.
• attention_mask -- List of indices specifying which tokens should be attended to by the model.
• labels -- List of token ids for tgt_texts.

The full set of keys [input_ids, attention_mask, labels], will only be returned if tgt_texts is passed. Otherwise, input_ids, attention_mask will be the only keys.

Source code in mindnlp\transformers\tokenization_utils_base.py

    def prepare_seq2seq_batch(
        self,
        src_texts: List[str],
        tgt_texts: Optional[List[str]] = None,
        max_length: Optional[int] = None,
        max_target_length: Optional[int] = None,
        padding: str = "longest",
        return_tensors: str = None,
        truncation: bool = True,
        **kwargs,
    ) -> BatchEncoding:
        """
        Prepare model inputs for translation. For best performance, translate one sentence at a time.

        Arguments:
            src_texts (`List[str]`):
                List of documents to summarize or source language texts.
            tgt_texts (`list`, *optional*):
                List of summaries or target language texts.
            max_length (`int`, *optional*):
                Controls the maximum length for encoder inputs (documents to summarize or source language texts) If
                left unset or set to `None`, this will use the predefined model maximum length if a maximum length is
                required by one of the truncation/padding parameters. If the model has no specific maximum input length
                (like XLNet) truncation/padding to a maximum length will be deactivated.
            max_target_length (`int`, *optional*):
                Controls the maximum length of decoder inputs (target language texts or summaries) If left unset or set
                to `None`, this will use the max_length value.
            padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`):
                Activates and controls padding. Accepts the following values:

                - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
                  sequence if provided).
                - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum
                  acceptable input length for the model if that argument is not provided.
                - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different
                  lengths).
            return_tensors (`str` or [`~utils.TensorType`], *optional*):
                If set, will return tensors instead of list of python integers. Acceptable values are:

                - `'tf'`: Return TensorFlow `tf.constant` objects.
                - `'pt'`: Return PyTorch `mindspore.Tensor` objects.
                - `'np'`: Return Numpy `np.ndarray` objects.
            truncation (`bool`, `str` or [`~tokenization_utils_base.TruncationStrategy`], *optional*, defaults to `True`):
                Activates and controls truncation. Accepts the following values:

                - `True` or `'longest_first'`: Truncate to a maximum length specified with the argument `max_length` or
                  to the maximum acceptable input length for the model if that argument is not provided. This will
                  truncate token by token, removing a token from the longest sequence in the pair if a pair of
                  sequences (or a batch of pairs) is provided.
                - `'only_first'`: Truncate to a maximum length specified with the argument `max_length` or to the
                  maximum acceptable input length for the model if that argument is not provided. This will only
                  truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
                - `'only_second'`: Truncate to a maximum length specified with the argument `max_length` or to the
                  maximum acceptable input length for the model if that argument is not provided. This will only
                  truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
                - `False` or `'do_not_truncate'` (default): No truncation (i.e., can output batch with sequence lengths
                  greater than the model maximum admissible input size).
            **kwargs:
                Additional keyword arguments passed along to `self.__call__`.

        Return:
            [`BatchEncoding`]: A [`BatchEncoding`] with the following fields:

            - **input_ids** -- List of token ids to be fed to the encoder.
            - **attention_mask** -- List of indices specifying which tokens should be attended to by the model.
            - **labels** -- List of token ids for tgt_texts.

            The full set of keys `[input_ids, attention_mask, labels]`, will only be returned if tgt_texts is passed.
            Otherwise, input_ids, attention_mask will be the only keys.
        """
        # docstyle-ignore
        formatted_warning = """
`prepare_seq2seq_batch` is deprecated. Use the regular
`__call__` method to prepare your inputs and targets.

Here is a short example:

model_inputs = tokenizer(src_texts, text_target=tgt_texts, ...)

If you either need to use different keyword arguments for the source and target texts, you should do two calls like
this:

model_inputs = tokenizer(src_texts, ...)
labels = tokenizer(text_target=tgt_texts, ...)
model_inputs["labels"] = labels["input_ids"]

See the documentation of your specific tokenizer for more details on the specific arguments to the tokenizer of choice.
For a more complete example, see the implementation of `prepare_seq2seq_batch`.
"""
        warnings.warn(formatted_warning, FutureWarning)
        # mBART-specific kwargs that should be ignored by other models.
        kwargs.pop("src_lang", None)
        kwargs.pop("tgt_lang", None)
        if max_length is None:
            max_length = self.model_max_length
        model_inputs = self(
            src_texts,
            add_special_tokens=True,
            return_tensors=return_tensors,
            max_length=max_length,
            padding=padding,
            truncation=truncation,
            **kwargs,
        )
        if tgt_texts is None:
            return model_inputs
        # Process tgt_texts
        if max_target_length is None:
            max_target_length = max_length
        with self.as_target_tokenizer():
            labels = self(
                tgt_texts,
                add_special_tokens=True,
                return_tensors=return_tensors,
                padding=padding,
                max_length=max_target_length,
                truncation=truncation,
                **kwargs,
            )
        model_inputs["labels"] = labels["input_ids"]
        return model_inputs

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.register_for_auto_class(auto_class='AutoTokenizer') classmethod

Register this class with a given auto class. This should only be used for custom tokenizers as the ones in the library are already mapped with AutoTokenizer.

This API is experimental and may have some slight breaking changes in the next releases.

PARAMETER	DESCRIPTION
auto_class	The auto class to register this new tokenizer with. TYPE: `str` or `type`, *optional*, defaults to `"AutoTokenizer"` DEFAULT: 'AutoTokenizer'

Source code in mindnlp\transformers\tokenization_utils_base.py

@classmethod
def register_for_auto_class(cls, auto_class="AutoTokenizer"):
    """
    Register this class with a given auto class. This should only be used for custom tokenizers as the ones in the
    library are already mapped with `AutoTokenizer`.

    <Tip warning={true}>

    This API is experimental and may have some slight breaking changes in the next releases.

    </Tip>

    Args:
        auto_class (`str` or `type`, *optional*, defaults to `"AutoTokenizer"`):
            The auto class to register this new tokenizer with.
    """
    if not isinstance(auto_class, str):
        auto_class = auto_class.__name__

    import mindnlp.transformers.models.auto as auto_module

    if not hasattr(auto_module, auto_class):
        raise ValueError(f"{auto_class} is not a valid auto class.")

    cls._auto_class = auto_class

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.save_pretrained(save_directory, legacy_format=None, filename_prefix=None, push_to_hub=False, **kwargs)

Save the full tokenizer state.

This method make sure the full tokenizer can then be re-loaded using the [~tokenization_utils_base.PreTrainedTokenizer.from_pretrained] class method..

Warning,None This won't save modifications you may have applied to the tokenizer after the instantiation (for instance, modifying tokenizer.do_lower_case after creation).

PARAMETER	DESCRIPTION
save_directory	The path to a directory where the tokenizer will be saved. TYPE: `str` or `os.PathLike`
legacy_format	Only applicable for a fast tokenizer. If unset (default), will save the tokenizer in the unified JSON format as well as in legacy format if it exists, i.e. with tokenizer specific vocabulary and a separate added_tokens files. If False, will only save the tokenizer in the unified JSON format. This format is incompatible with "slow" tokenizers (not powered by the tokenizers library), so the tokenizer will not be able to be loaded in the corresponding "slow" tokenizer. If True, will save the tokenizer in legacy format. If the "slow" tokenizer doesn't exits, a value error is raised. TYPE: `bool`, *optional* DEFAULT: None
filename_prefix	A prefix to add to the names of the files saved by the tokenizer. TYPE: `str`, *optional* DEFAULT: None
push_to_hub	Whether or not to push your model to the Hugging Face model hub after saving it. You can specify the repository you want to push to with repo_id (will default to the name of save_directory in your namespace). TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False
kwargs	Additional key word arguments passed along to the [~utils.PushToHubMixin.push_to_hub] method. TYPE: `Dict[str, Any]`, *optional* DEFAULT: {}

RETURNS	DESCRIPTION
Tuple[str]	A tuple of str: The files saved.

Source code in mindnlp\transformers\tokenization_utils_base.py

def save_pretrained(
    self,
    save_directory: Union[str, os.PathLike],
    legacy_format: Optional[bool] = None,
    filename_prefix: Optional[str] = None,
    push_to_hub: bool = False,
    **kwargs,
) -> Tuple[str]:
    """
    Save the full tokenizer state.


    This method make sure the full tokenizer can then be re-loaded using the
    [`~tokenization_utils_base.PreTrainedTokenizer.from_pretrained`] class method..

    Warning,None This won't save modifications you may have applied to the tokenizer after the instantiation (for
    instance, modifying `tokenizer.do_lower_case` after creation).

    Args:
        save_directory (`str` or `os.PathLike`): The path to a directory where the tokenizer will be saved.
        legacy_format (`bool`, *optional*):
            Only applicable for a fast tokenizer. If unset (default), will save the tokenizer in the unified JSON
            format as well as in legacy format if it exists, i.e. with tokenizer specific vocabulary and a separate
            added_tokens files.

            If `False`, will only save the tokenizer in the unified JSON format. This format is incompatible with
            "slow" tokenizers (not powered by the *tokenizers* library), so the tokenizer will not be able to be
            loaded in the corresponding "slow" tokenizer.

            If `True`, will save the tokenizer in legacy format. If the "slow" tokenizer doesn't exits, a value
            error is raised.
        filename_prefix (`str`, *optional*):
            A prefix to add to the names of the files saved by the tokenizer.
        push_to_hub (`bool`, *optional*, defaults to `False`):
            Whether or not to push your model to the Hugging Face model hub after saving it. You can specify the
            repository you want to push to with `repo_id` (will default to the name of `save_directory` in your
            namespace).
        kwargs (`Dict[str, Any]`, *optional*):
            Additional key word arguments passed along to the [`~utils.PushToHubMixin.push_to_hub`] method.

    Returns:
        A tuple of `str`: The files saved.
    """
    use_auth_token = kwargs.pop("use_auth_token", None)

    if use_auth_token is not None:
        warnings.warn(
            "The `use_auth_token` argument is deprecated. Please use `token` instead.",
            FutureWarning,
        )
        if kwargs.get("token", None) is not None:
            raise ValueError(
                "`token` and `use_auth_token` are both specified. Please set only the argument `token`."
            )
        kwargs["token"] = use_auth_token

    if os.path.isfile(save_directory):
        logger.error(f"Provided path ({save_directory}) should be a directory, not a file")
        return

    os.makedirs(save_directory, exist_ok=True)

    if push_to_hub:
        commit_message = kwargs.pop("commit_message", None)
        repo_id = kwargs.pop("repo_id", save_directory.split(os.path.sep)[-1])
        repo_id = self._create_repo(repo_id, **kwargs)
        files_timestamps = self._get_files_timestamps(save_directory)

    special_tokens_map_file = os.path.join(
        save_directory, (filename_prefix + "-" if filename_prefix else "") + SPECIAL_TOKENS_MAP_FILE
    )
    tokenizer_config_file = os.path.join(
        save_directory, (filename_prefix + "-" if filename_prefix else "") + TOKENIZER_CONFIG_FILE
    )

    tokenizer_config = copy.deepcopy(self.init_kwargs)

    # Let's save the init kwargs
    target_keys = set(self.init_kwargs.keys())
    # Let's save the special tokens map (only the strings)
    target_keys.update(["model_max_length", "clean_up_tokenization_spaces"])

    for k in target_keys:
        if hasattr(self, k):
            tokenizer_config[k] = getattr(self, k)

    # Let's make sure we properly save the special tokens.
    tokenizer_config.update(self.special_tokens_map)

    if self.chat_template is not None:
        if isinstance(self.chat_template, dict):
            # Chat template dicts are saved to the config as lists of dicts with fixed key names.
            # They will be reconstructed as a single dict during loading.
            tokenizer_config["chat_template"] = [{"name": k, "template": v} for k, v in self.chat_template.items()]
        else:
            tokenizer_config["chat_template"] = self.chat_template

    if len(self.init_inputs) > 0:
        tokenizer_config["init_inputs"] = copy.deepcopy(self.init_inputs)
    for file_id in self.vocab_files_names.keys():
        tokenizer_config.pop(file_id, None)

    # no typefields, this way old fast and slow can load it
    tokenizer_config = self.convert_added_tokens(tokenizer_config, add_type_field=True, save=True)

    # Process added tokens seperatly: allows previous versions to ignore it!
    added_tokens = {}
    for key, value in self.added_tokens_decoder.items():
        added_tokens[key] = value.__getstate__()
    tokenizer_config["added_tokens_decoder"] = added_tokens

    # Add tokenizer class to the tokenizer config to be able to reload it with from_pretrained
    tokenizer_class = self.__class__.__name__
    # Remove the Fast at the end unless we have a special `PreTrainedTokenizerFast`
    if tokenizer_class.endswith("Fast") and tokenizer_class != "PreTrainedTokenizerFast":
        tokenizer_class = tokenizer_class[:-4]
    tokenizer_config["tokenizer_class"] = tokenizer_class
    if getattr(self, "_auto_map", None) is not None:
        tokenizer_config["auto_map"] = self._auto_map
    if getattr(self, "_processor_class", None) is not None:
        tokenizer_config["processor_class"] = self._processor_class

    # remove private information
    if "name_or_path" in tokenizer_config:
        tokenizer_config.pop("name_or_path")
        tokenizer_config.pop("special_tokens_map_file", None)
        tokenizer_config.pop("tokenizer_file", None)
    if "device_map" in tokenizer_config:
        tokenizer_config.pop("device_map")

    with open(tokenizer_config_file, "w", encoding="utf-8") as f:
        out_str = json.dumps(tokenizer_config, indent=2, sort_keys=True, ensure_ascii=False) + "\n"
        f.write(out_str)
    logger.info(f"tokenizer config file saved in {tokenizer_config_file}")

    # Sanitize AddedTokens in special_tokens_map

    # kept for forward compatibility, will be removed in transoformers 5. Typefields are not saved for FC, special should not be save either
    write_dict = self.convert_added_tokens(self.special_tokens_map_extended, save=True, add_type_field=False)
    with open(special_tokens_map_file, "w", encoding="utf-8") as f:
        out_str = json.dumps(write_dict, indent=2, sort_keys=True, ensure_ascii=False) + "\n"
        f.write(out_str)
    logger.info(f"Special tokens file saved in {special_tokens_map_file}")

    file_names = (tokenizer_config_file, special_tokens_map_file)

    save_files = self._save_pretrained(
        save_directory=save_directory,
        file_names=file_names,
        legacy_format=legacy_format,
        filename_prefix=filename_prefix,
    )

    if push_to_hub:
        self._upload_modified_files(
            save_directory,
            repo_id,
            files_timestamps,
            commit_message=commit_message,
            token=kwargs.get("token"),
        )

    return save_files

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.save_vocabulary(save_directory, filename_prefix=None)

Save only the vocabulary of the tokenizer (vocabulary + added tokens).

This method won't save the configuration and special token mappings of the tokenizer. Use [~PreTrainedTokenizerFast._save_pretrained] to save the whole state of the tokenizer.

PARAMETER	DESCRIPTION
save_directory	The directory in which to save the vocabulary. TYPE: `str`
filename_prefix	An optional prefix to add to the named of the saved files. TYPE: `str`, *optional* DEFAULT: None

RETURNS	DESCRIPTION
Tuple[str]	Tuple(str): Paths to the files saved.

Source code in mindnlp\transformers\tokenization_utils_base.py

def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
    """
    Save only the vocabulary of the tokenizer (vocabulary + added tokens).

    This method won't save the configuration and special token mappings of the tokenizer. Use
    [`~PreTrainedTokenizerFast._save_pretrained`] to save the whole state of the tokenizer.

    Args:
        save_directory (`str`):
            The directory in which to save the vocabulary.
        filename_prefix (`str`, *optional*):
            An optional prefix to add to the named of the saved files.

    Returns:
        `Tuple(str)`: Paths to the files saved.
    """
    raise NotImplementedError

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.tokenize(text, pair=None, add_special_tokens=False, **kwargs)

Converts a string into a sequence of tokens, replacing unknown tokens with the unk_token.

PARAMETER	DESCRIPTION
text	The sequence to be encoded. TYPE: `str`
pair	A second sequence to be encoded with the first. TYPE: `str`, *optional* DEFAULT: None
add_special_tokens	Whether or not to add the special tokens associated with the corresponding model. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False
kwargs	Will be passed to the underlying model specific encode method. See details in [~PreTrainedTokenizerBase.__call__] TYPE: additional keyword arguments, *optional* DEFAULT: {}

RETURNS	DESCRIPTION
List[str]	List[str]: The list of tokens.

Source code in mindnlp\transformers\tokenization_utils_base.py

def tokenize(self, text: str, pair: Optional[str] = None, add_special_tokens: bool = False, **kwargs) -> List[str]:
    """
    Converts a string into a sequence of tokens, replacing unknown tokens with the `unk_token`.

    Args:
        text (`str`):
            The sequence to be encoded.
        pair (`str`, *optional*):
            A second sequence to be encoded with the first.
        add_special_tokens (`bool`, *optional*, defaults to `False`):
            Whether or not to add the special tokens associated with the corresponding model.
        kwargs (additional keyword arguments, *optional*):
            Will be passed to the underlying model specific encode method. See details in
            [`~PreTrainedTokenizerBase.__call__`]

    Returns:
        `List[str]`: The list of tokens.
    """
    raise NotImplementedError

mindnlp.transformers.tokenization_utils_base.PreTrainedTokenizerBase.truncate_sequences(ids, pair_ids=None, num_tokens_to_remove=0, truncation_strategy='longest_first', stride=0)

Truncates a sequence pair in-place following the strategy.

PARAMETER	DESCRIPTION
ids	Tokenized input ids of the first sequence. Can be obtained from a string by chaining the tokenize and convert_tokens_to_ids methods. TYPE: `List[int]`
pair_ids	Tokenized input ids of the second sequence. Can be obtained from a string by chaining the tokenize and convert_tokens_to_ids methods. TYPE: `List[int]`, *optional* DEFAULT: None
num_tokens_to_remove	Number of tokens to remove using the truncation strategy. TYPE: `int`, *optional*, defaults to 0 DEFAULT: 0
truncation_strategy	The strategy to follow for truncation. Can be: 'longest_first': Truncate to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. This will truncate token by token, removing a token from the longest sequence in the pair if a pair of sequences (or a batch of pairs) is provided. 'only_first': Truncate to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. This will only truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided. 'only_second': Truncate to a maximum length specified with the argument max_length or to the maximum acceptable input length for the model if that argument is not provided. This will only truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided. 'do_not_truncate' (default): No truncation (i.e., can output batch with sequence lengths greater than the model maximum admissible input size). TYPE: `str` or [`~tokenization_utils_base.TruncationStrategy`], *optional*, defaults to `'longest_first'` DEFAULT: 'longest_first'
stride	If set to a positive number, the overflowing tokens returned will contain some tokens from the main sequence returned. The value of this argument defines the number of additional tokens. TYPE: `int`, *optional*, defaults to 0 DEFAULT: 0

RETURNS	DESCRIPTION
List[int]	Tuple[List[int], List[int], List[int]]: The truncated ids, the truncated pair_ids and the list of
List[int]	overflowing tokens. Note: The longest_first strategy returns empty list of overflowing tokens if a pair
List[int]	of sequences (or a batch of pairs) is provided.

Source code in mindnlp\transformers\tokenization_utils_base.py

def truncate_sequences(
    self,
    ids: List[int],
    pair_ids: Optional[List[int]] = None,
    num_tokens_to_remove: int = 0,
    truncation_strategy: Union[str, TruncationStrategy] = "longest_first",
    stride: int = 0,
) -> Tuple[List[int], List[int], List[int]]:
    """
    Truncates a sequence pair in-place following the strategy.

    Args:
        ids (`List[int]`):
            Tokenized input ids of the first sequence. Can be obtained from a string by chaining the `tokenize` and
            `convert_tokens_to_ids` methods.
        pair_ids (`List[int]`, *optional*):
            Tokenized input ids of the second sequence. Can be obtained from a string by chaining the `tokenize`
            and `convert_tokens_to_ids` methods.
        num_tokens_to_remove (`int`, *optional*, defaults to 0):
            Number of tokens to remove using the truncation strategy.
        truncation_strategy (`str` or [`~tokenization_utils_base.TruncationStrategy`], *optional*, defaults to `'longest_first'`):
            The strategy to follow for truncation. Can be:

            - `'longest_first'`: Truncate to a maximum length specified with the argument `max_length` or to the
              maximum acceptable input length for the model if that argument is not provided. This will truncate
              token by token, removing a token from the longest sequence in the pair if a pair of sequences (or a
              batch of pairs) is provided.
            - `'only_first'`: Truncate to a maximum length specified with the argument `max_length` or to the
              maximum acceptable input length for the model if that argument is not provided. This will only
              truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
            - `'only_second'`: Truncate to a maximum length specified with the argument `max_length` or to the
              maximum acceptable input length for the model if that argument is not provided. This will only
              truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
            - `'do_not_truncate'` (default): No truncation (i.e., can output batch with sequence lengths greater
              than the model maximum admissible input size).
        stride (`int`, *optional*, defaults to 0):
            If set to a positive number, the overflowing tokens returned will contain some tokens from the main
            sequence returned. The value of this argument defines the number of additional tokens.

    Returns:
        `Tuple[List[int], List[int], List[int]]`: The truncated `ids`, the truncated `pair_ids` and the list of
        overflowing tokens. Note: The *longest_first* strategy returns empty list of overflowing tokens if a pair
        of sequences (or a batch of pairs) is provided.
    """
    if num_tokens_to_remove <= 0:
        return ids, pair_ids, []

    if not isinstance(truncation_strategy, TruncationStrategy):
        truncation_strategy = TruncationStrategy(truncation_strategy)

    overflowing_tokens = []
    if truncation_strategy == TruncationStrategy.ONLY_FIRST or (
        truncation_strategy == TruncationStrategy.LONGEST_FIRST and pair_ids is None
    ):
        if len(ids) > num_tokens_to_remove:
            window_len = min(len(ids), stride + num_tokens_to_remove)
            if self.truncation_side == "left":
                overflowing_tokens = ids[:window_len]
                ids = ids[num_tokens_to_remove:]
            elif self.truncation_side == "right":
                overflowing_tokens = ids[-window_len:]
                ids = ids[:-num_tokens_to_remove]
            else:
                raise ValueError(f"invalid truncation strategy: {self.truncation_side}, use 'left' or 'right'.")

        else:
            error_msg = (
                f"We need to remove {num_tokens_to_remove} to truncate the input "
                f"but the first sequence has a length {len(ids)}. "
            )
            if truncation_strategy == TruncationStrategy.ONLY_FIRST:
                error_msg = (
                    error_msg + "Please select another truncation strategy than "
                    f"{truncation_strategy}, for instance 'longest_first' or 'only_second'."
                )
            logger.error(error_msg)
    elif truncation_strategy == TruncationStrategy.LONGEST_FIRST:
        logger.warning(
            "Be aware, overflowing tokens are not returned for the setting you have chosen,"
            f" i.e. sequence pairs with the '{TruncationStrategy.LONGEST_FIRST.value}' "
            "truncation strategy. So the returned list will always be empty even if some "
            "tokens have been removed."
        )
        len_pair_ids = len(pair_ids) if pair_ids is not None else 0
        len_ids = len(ids)
        first_remove = min(abs(len_pair_ids - len_ids), num_tokens_to_remove)
        second_remove = num_tokens_to_remove - first_remove
        if len_ids > len_pair_ids:
            ids_to_move = first_remove + second_remove // 2
            pair_ids_to_move = second_remove - second_remove // 2
        else:
            ids_to_move = second_remove // 2
            pair_ids_to_move = first_remove + second_remove - (second_remove // 2)

        if self.truncation_side == "right":
            ids = ids[:-ids_to_move] if ids_to_move > 0 else ids
            pair_ids = pair_ids[:-pair_ids_to_move] if pair_ids is not None and pair_ids_to_move > 0 else pair_ids
        elif self.truncation_side == "left":
            ids = ids[ids_to_move:]
            pair_ids = pair_ids[pair_ids_to_move:] if pair_ids is not None else None
        else:
            raise ValueError(f"invalid truncation strategy:{self.truncation_side}")

    elif truncation_strategy == TruncationStrategy.ONLY_SECOND and pair_ids is not None:
        if len(pair_ids) > num_tokens_to_remove:
            window_len = min(len(pair_ids), stride + num_tokens_to_remove)
            if self.truncation_side == "right":
                overflowing_tokens = pair_ids[-window_len:]
                pair_ids = pair_ids[:-num_tokens_to_remove]
            elif self.truncation_side == "left":
                overflowing_tokens = pair_ids[:window_len]
                pair_ids = pair_ids[num_tokens_to_remove:]
            else:
                raise ValueError(f"invalid truncation strategy:{self.truncation_side}")
        else:
            logger.error(
                f"We need to remove {num_tokens_to_remove} to truncate the input "
                f"but the second sequence has a length {len(pair_ids)}. "
                f"Please select another truncation strategy than {truncation_strategy}, "
                "for instance 'longest_first' or 'only_first'."
            )

    return (ids, pair_ids, overflowing_tokens)

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin

A mixin derived by [PreTrainedTokenizer] and [PreTrainedTokenizerFast] to handle specific behaviors related to special tokens. In particular, this class hold the attributes which can be used to directly access these special tokens in a model-independent manner and allow to set and update the special tokens.

PARAMETER	DESCRIPTION
bos_token	A special token representing the beginning of a sentence. TYPE: `str` or `tokenizers.AddedToken`, *optional*
eos_token	A special token representing the end of a sentence. TYPE: `str` or `tokenizers.AddedToken`, *optional*
unk_token	A special token representing an out-of-vocabulary token. TYPE: `str` or `tokenizers.AddedToken`, *optional*
sep_token	A special token separating two different sentences in the same input (used by BERT for instance). TYPE: `str` or `tokenizers.AddedToken`, *optional*
pad_token	A special token used to make arrays of tokens the same size for batching purpose. Will then be ignored by attention mechanisms or loss computation. TYPE: `str` or `tokenizers.AddedToken`, *optional*
cls_token	A special token representing the class of the input (used by BERT for instance). TYPE: `str` or `tokenizers.AddedToken`, *optional*
mask_token	A special token representing a masked token (used by masked-language modeling pretraining objectives, like BERT). TYPE: `str` or `tokenizers.AddedToken`, *optional*
additional_special_tokens	A tuple or a list of additional tokens, which will be marked as special, meaning that they will be skipped when decoding if skip_special_tokens is set to True. TYPE: tuple or list of `str` or `tokenizers.AddedToken`, *optional*

Source code in mindnlp\transformers\tokenization_utils_base.py

class SpecialTokensMixin:
    """
    A mixin derived by [`PreTrainedTokenizer`] and [`PreTrainedTokenizerFast`] to handle specific behaviors related to
    special tokens. In particular, this class hold the attributes which can be used to directly access these special
    tokens in a model-independent manner and allow to set and update the special tokens.

    Args:
        bos_token (`str` or `tokenizers.AddedToken`, *optional*):
            A special token representing the beginning of a sentence.
        eos_token (`str` or `tokenizers.AddedToken`, *optional*):
            A special token representing the end of a sentence.
        unk_token (`str` or `tokenizers.AddedToken`, *optional*):
            A special token representing an out-of-vocabulary token.
        sep_token (`str` or `tokenizers.AddedToken`, *optional*):
            A special token separating two different sentences in the same input (used by BERT for instance).
        pad_token (`str` or `tokenizers.AddedToken`, *optional*):
            A special token used to make arrays of tokens the same size for batching purpose. Will then be ignored by
            attention mechanisms or loss computation.
        cls_token (`str` or `tokenizers.AddedToken`, *optional*):
            A special token representing the class of the input (used by BERT for instance).
        mask_token (`str` or `tokenizers.AddedToken`, *optional*):
            A special token representing a masked token (used by masked-language modeling pretraining objectives, like
            BERT).
        additional_special_tokens (tuple or list of `str` or `tokenizers.AddedToken`, *optional*):
            A tuple or a list of additional tokens, which will be marked as `special`, meaning that they will be
            skipped when decoding if `skip_special_tokens` is set to `True`.
    """

    SPECIAL_TOKENS_ATTRIBUTES = [
        "bos_token",
        "eos_token",
        "unk_token",
        "sep_token",
        "pad_token",
        "cls_token",
        "mask_token",
        "additional_special_tokens",
    ]

    def __init__(self, verbose=False, **kwargs):
        self._bos_token = None
        self._eos_token = None
        self._unk_token = None
        self._sep_token = None
        self._pad_token = None
        self._cls_token = None
        self._mask_token = None
        self._pad_token_type_id = 0
        self._additional_special_tokens = []
        self.verbose = verbose

        # We directly set the hidden value to allow initialization with special tokens
        # which are not yet in the vocabulary. Necessary for serialization/de-serialization
        # TODO clean this up at some point (probably by switching to fast tokenizers)

        for key, value in kwargs.items():
            if value is None:
                continue
            if key in self.SPECIAL_TOKENS_ATTRIBUTES:
                if key == "additional_special_tokens":
                    assert isinstance(value, (list, tuple)), f"Value {value} is not a list or tuple"
                    assert all(
                        isinstance(t, (str, AddedToken)) for t in value
                    ), "One of the tokens is not a string or an AddedToken"
                    setattr(self, key, value)
                elif isinstance(value, (str, AddedToken)):
                    setattr(self, key, value)
                else:
                    raise TypeError(f"Special token {key} has to be either str or AddedToken but got: {type(value)}")

    def sanitize_special_tokens(self) -> int:
        """
        The `sanitize_special_tokens` is now deprecated kept for backward compatibility and will be removed in
        transformers v5.
        """
        logger.warning_once("The `sanitize_special_tokens` will be removed in transformers v5.")
        return self.add_tokens(self.all_special_tokens_extended, special_tokens=True)

    def add_special_tokens(
        self, special_tokens_dict: Dict[str, Union[str, AddedToken]], replace_additional_special_tokens=True
    ) -> int:
        """
        Add a dictionary of special tokens (eos, pad, cls, etc.) to the encoder and link them to class attributes. If
        special tokens are NOT in the vocabulary, they are added to it (indexed starting from the last index of the
        current vocabulary).

        When adding new tokens to the vocabulary, you should make sure to also resize the token embedding matrix of the
        model so that its embedding matrix matches the tokenizer.

        In order to do that, please use the [`~PreTrainedModel.resize_token_embeddings`] method.

        Using `add_special_tokens` will ensure your special tokens can be used in several ways:

        - Special tokens can be skipped when decoding using `skip_special_tokens = True`.
        - Special tokens are carefully handled by the tokenizer (they are never split), similar to `AddedTokens`.
        - You can easily refer to special tokens using tokenizer class attributes like `tokenizer.cls_token`. This
          makes it easy to develop model-agnostic training and fine-tuning scripts.

        When possible, special tokens are already registered for provided pretrained models (for instance
        [`BertTokenizer`] `cls_token` is already registered to be :obj*'[CLS]'* and XLM's one is also registered to be
        `'</s>'`).

        Args:
            special_tokens_dict (dictionary *str* to *str* or `tokenizers.AddedToken`):
                Keys should be in the list of predefined special attributes: [`bos_token`, `eos_token`, `unk_token`,
                `sep_token`, `pad_token`, `cls_token`, `mask_token`, `additional_special_tokens`].

                Tokens are only added if they are not already in the vocabulary (tested by checking if the tokenizer
                assign the index of the `unk_token` to them).
            replace_additional_special_tokens (`bool`, *optional*,, defaults to `True`):
                If `True`, the existing list of additional special tokens will be replaced by the list provided in
                `special_tokens_dict`. Otherwise, `self._additional_special_tokens` is just extended. In the former
                case, the tokens will NOT be removed from the tokenizer's full vocabulary - they are only being flagged
                as non-special tokens. Remember, this only affects which tokens are skipped during decoding, not the
                `added_tokens_encoder` and `added_tokens_decoder`. This means that the previous
                `additional_special_tokens` are still added tokens, and will not be split by the model.

        Returns:
            `int`: Number of tokens added to the vocabulary.

        Examples:

        ```python
        # Let's see how to add a new classification token to GPT-2
        tokenizer = GPT2Tokenizer.from_pretrained("openai-community/gpt2")
        model = GPT2Model.from_pretrained("openai-community/gpt2")

        special_tokens_dict = {"cls_token": "<CLS>"}

        num_added_toks = tokenizer.add_special_tokens(special_tokens_dict)
        print("We have added", num_added_toks, "tokens")
        # Notice: resize_token_embeddings expect to receive the full size of the new vocabulary, i.e., the length of the tokenizer.
        model.resize_token_embeddings(len(tokenizer))

        assert tokenizer.cls_token == "<CLS>"
        ```"""
        if not special_tokens_dict:
            return 0

        added_tokens = []
        for key, value in special_tokens_dict.items():
            assert key in self.SPECIAL_TOKENS_ATTRIBUTES, f"Key {key} is not a special token"

            if self.verbose:
                logger.info(f"Assigning {value} to the {key} key of the tokenizer")

            if key == "additional_special_tokens":
                assert isinstance(value, (list, tuple)) and all(
                    isinstance(t, (str, AddedToken)) for t in value
                ), f"Tokens {value} for key {key} should all be str or AddedToken instances"

                to_add = []
                for token in value:
                    if isinstance(token, str):
                        # for legacy purpose we default to stripping. `test_add_tokens_tokenizer` depends on this
                        token = AddedToken(token, rstrip=False, lstrip=False, normalized=False, special=True)
                    if not replace_additional_special_tokens and str(token) in self.additional_special_tokens:
                        continue
                    to_add.append(token)
                if replace_additional_special_tokens and len(to_add) > 0:
                    setattr(self, key, list(to_add))
                else:
                    self._additional_special_tokens.extend(to_add)
                added_tokens += to_add

            else:
                if not isinstance(value, (str, AddedToken)):
                    raise ValueError(f"Token {value} for key {key} should be a str or an AddedToken instance")
                if isinstance(value, (str)):
                    # for legacy purpose we default to stripping. `False` depends on this
                    value = AddedToken(value, rstrip=False, lstrip=False, normalized=False, special=True)
                if isinstance(value, AddedToken):
                    setattr(self, key, value)
                if value not in added_tokens:
                    added_tokens.append(value)

        # if we are adding tokens that were not part of the vocab, we ought to add them
        added_tokens = self.add_tokens(added_tokens, special_tokens=True)
        return added_tokens

    def add_tokens(
        self, new_tokens: Union[str, AddedToken, List[Union[str, AddedToken]]], special_tokens: bool = False
    ) -> int:
        """
        Add a list of new tokens to the tokenizer class. If the new tokens are not in the vocabulary, they are added to
        it with indices starting from length of the current vocabulary and and will be isolated before the tokenization
        algorithm is applied. Added tokens and tokens from the vocabulary of the tokenization algorithm are therefore
        not treated in the same way.

        Note, when adding new tokens to the vocabulary, you should make sure to also resize the token embedding matrix
        of the model so that its embedding matrix matches the tokenizer.

        In order to do that, please use the [`~PreTrainedModel.resize_token_embeddings`] method.

        Args:
            new_tokens (`str`, `tokenizers.AddedToken` or a list of *str* or `tokenizers.AddedToken`):
                Tokens are only added if they are not already in the vocabulary. `tokenizers.AddedToken` wraps a string
                token to let you personalize its behavior: whether this token should only match against a single word,
                whether this token should strip all potential whitespaces on the left side, whether this token should
                strip all potential whitespaces on the right side, etc.
            special_tokens (`bool`, *optional*, defaults to `False`):
                Can be used to specify if the token is a special token. This mostly change the normalization behavior
                (special tokens like CLS or [MASK] are usually not lower-cased for instance).

                See details for `tokenizers.AddedToken` in HuggingFace tokenizers library.

        Returns:
            `int`: Number of tokens added to the vocabulary.

        Examples:

        ```python
        # Let's see how to increase the vocabulary of Bert model and tokenizer
        tokenizer = BertTokenizerFast.from_pretrained("google-bert/bert-base-uncased")
        model = BertModel.from_pretrained("google-bert/bert-base-uncased")

        num_added_toks = tokenizer.add_tokens(["new_tok1", "my_new-tok2"])
        print("We have added", num_added_toks, "tokens")
        # Notice: resize_token_embeddings expect to receive the full size of the new vocabulary, i.e., the length of the tokenizer.
        model.resize_token_embeddings(len(tokenizer))
        ```"""
        if not new_tokens:
            return 0

        if not isinstance(new_tokens, (list, tuple)):
            new_tokens = [new_tokens]

        return self._add_tokens(new_tokens, special_tokens=special_tokens)

    def _add_tokens(self, new_tokens: Union[List[str], List[AddedToken]], special_tokens: bool = False) -> int:
        raise NotImplementedError

    @property
    def bos_token(self) -> str:
        """
        `str`: Beginning of sentence token. Log an error if used while not having been set.
        """
        if self._bos_token is None:
            if self.verbose:
                logger.error("Using bos_token, but it is not set yet.")
            return None
        return str(self._bos_token)

    @property
    def eos_token(self) -> str:
        """
        `str`: End of sentence token. Log an error if used while not having been set.
        """
        if self._eos_token is None:
            if self.verbose:
                logger.error("Using eos_token, but it is not set yet.")
            return None
        return str(self._eos_token)

    @property
    def unk_token(self) -> str:
        """
        `str`: Unknown token. Log an error if used while not having been set.
        """
        if self._unk_token is None:
            if self.verbose:
                logger.error("Using unk_token, but it is not set yet.")
            return None
        return str(self._unk_token)

    @property
    def sep_token(self) -> str:
        """
        `str`: Separation token, to separate context and query in an input sequence. Log an error if used while not
        having been set.
        """
        if self._sep_token is None:
            if self.verbose:
                logger.error("Using sep_token, but it is not set yet.")
            return None
        return str(self._sep_token)

    @property
    def pad_token(self) -> str:
        """
        `str`: Padding token. Log an error if used while not having been set.
        """
        if self._pad_token is None:
            if self.verbose:
                logger.error("Using pad_token, but it is not set yet.")
            return None
        return str(self._pad_token)

    @property
    def cls_token(self) -> str:
        """
        `str`: Classification token, to extract a summary of an input sequence leveraging self-attention along the full
        depth of the model. Log an error if used while not having been set.
        """
        if self._cls_token is None:
            if self.verbose:
                logger.error("Using cls_token, but it is not set yet.")
            return None
        return str(self._cls_token)

    @property
    def mask_token(self) -> str:
        """
        `str`: Mask token, to use when training a model with masked-language modeling. Log an error if used while not
        having been set.
        """
        if self._mask_token is None:
            if self.verbose:
                logger.error("Using mask_token, but it is not set yet.")
            return None
        return str(self._mask_token)

    @property
    def additional_special_tokens(self) -> List[str]:
        """
        `List[str]`: All the additional special tokens you may want to use. Log an error if used while not having been
        set.
        """
        if self._additional_special_tokens is None:
            if self.verbose:
                logger.error("Using additional_special_tokens, but it is not set yet.")
            return None
        return [str(tok) for tok in self._additional_special_tokens]

    @bos_token.setter
    def bos_token(self, value):
        if not isinstance(value, (str, AddedToken)) and value is not None:
            raise ValueError("Cannot set a non-string value as the BOS token")
        self._bos_token = value

    @eos_token.setter
    def eos_token(self, value):
        if not isinstance(value, (str, AddedToken)) and value is not None:
            raise ValueError("Cannot set a non-string value as the EOS token")
        self._eos_token = value

    @unk_token.setter
    def unk_token(self, value):
        if not isinstance(value, (str, AddedToken)) and value is not None:
            raise ValueError("Cannot set a non-string value as the UNK token")
        self._unk_token = value

    @sep_token.setter
    def sep_token(self, value):
        if not isinstance(value, (str, AddedToken)) and value is not None:
            raise ValueError("Cannot set a non-string value as the SEP token")
        self._sep_token = value

    @pad_token.setter
    def pad_token(self, value):
        if not isinstance(value, (str, AddedToken)) and value is not None:
            raise ValueError("Cannot set a non-string value as the PAD token")
        self._pad_token = value

    @cls_token.setter
    def cls_token(self, value):
        if not isinstance(value, (str, AddedToken)) and value is not None:
            raise ValueError("Cannot set a non-string value as the CLS token")
        self._cls_token = value

    @mask_token.setter
    def mask_token(self, value):
        if not isinstance(value, (str, AddedToken)) and value is not None:
            raise ValueError("Cannot set a non-string value as the MASK token")
        self._mask_token = value

    @additional_special_tokens.setter
    def additional_special_tokens(self, value):
        self._additional_special_tokens = value if value is not None else None

    @property
    def bos_token_id(self) -> Optional[int]:
        """
        `Optional[int]`: Id of the beginning of sentence token in the vocabulary. Returns `None` if the token has not
        been set.
        """
        if self._bos_token is None:
            return None
        return self.convert_tokens_to_ids(self.bos_token)

    @property
    def eos_token_id(self) -> Optional[int]:
        """
        `Optional[int]`: Id of the end of sentence token in the vocabulary. Returns `None` if the token has not been
        set.
        """
        if self._eos_token is None:
            return None
        return self.convert_tokens_to_ids(self.eos_token)

    @property
    def unk_token_id(self) -> Optional[int]:
        """
        `Optional[int]`: Id of the unknown token in the vocabulary. Returns `None` if the token has not been set.
        """
        if self._unk_token is None:
            return None
        return self.convert_tokens_to_ids(self.unk_token)

    @property
    def sep_token_id(self) -> Optional[int]:
        """
        `Optional[int]`: Id of the separation token in the vocabulary, to separate context and query in an input
        sequence. Returns `None` if the token has not been set.
        """
        if self._sep_token is None:
            return None
        return self.convert_tokens_to_ids(self.sep_token)

    @property
    def pad_token_id(self) -> Optional[int]:
        """
        `Optional[int]`: Id of the padding token in the vocabulary. Returns `None` if the token has not been set.
        """
        if self._pad_token is None:
            return None
        return self.convert_tokens_to_ids(self.pad_token)

    @property
    def pad_token_type_id(self) -> int:
        """
        `int`: Id of the padding token type in the vocabulary.
        """
        return self._pad_token_type_id

    @property
    def cls_token_id(self) -> Optional[int]:
        """
        `Optional[int]`: Id of the classification token in the vocabulary, to extract a summary of an input sequence
        leveraging self-attention along the full depth of the model.

        Returns `None` if the token has not been set.
        """
        if self._cls_token is None:
            return None
        return self.convert_tokens_to_ids(self.cls_token)

    @property
    def mask_token_id(self) -> Optional[int]:
        """
        `Optional[int]`: Id of the mask token in the vocabulary, used when training a model with masked-language
        modeling. Returns `None` if the token has not been set.
        """
        if self._mask_token is None:
            return None
        return self.convert_tokens_to_ids(self.mask_token)

    @property
    def additional_special_tokens_ids(self) -> List[int]:
        """
        `List[int]`: Ids of all the additional special tokens in the vocabulary. Log an error if used while not having
        been set.
        """
        return self.convert_tokens_to_ids(self.additional_special_tokens)

    @bos_token_id.setter
    def bos_token_id(self, value):
        self._bos_token = self.convert_ids_to_tokens(value) if value is not None else None

    @eos_token_id.setter
    def eos_token_id(self, value):
        self._eos_token = self.convert_ids_to_tokens(value) if value is not None else None

    @unk_token_id.setter
    def unk_token_id(self, value):
        self._unk_token = self.convert_ids_to_tokens(value) if value is not None else None

    @sep_token_id.setter
    def sep_token_id(self, value):
        self._sep_token = self.convert_ids_to_tokens(value) if value is not None else None

    @pad_token_id.setter
    def pad_token_id(self, value):
        self._pad_token = self.convert_ids_to_tokens(value) if value is not None else None

    @cls_token_id.setter
    def cls_token_id(self, value):
        self._cls_token = self.convert_ids_to_tokens(value) if value is not None else None

    @mask_token_id.setter
    def mask_token_id(self, value):
        self._mask_token = self.convert_ids_to_tokens(value) if value is not None else None

    @additional_special_tokens_ids.setter
    def additional_special_tokens_ids(self, values):
        self._additional_special_tokens = [self.convert_ids_to_tokens(value) for value in values]

    @property
    def special_tokens_map(self) -> Dict[str, Union[str, List[str]]]:
        """
        `Dict[str, Union[str, List[str]]]`: A dictionary mapping special token class attributes (`cls_token`,
        `unk_token`, etc.) to their values (`'<unk>'`, `'<cls>'`, etc.).

        Convert potential tokens of `tokenizers.AddedToken` type to string.
        """
        set_attr = {}
        for attr in self.SPECIAL_TOKENS_ATTRIBUTES:
            attr_value = getattr(self, attr)
            if attr_value:
                set_attr[attr] = attr_value
        return set_attr

    @property
    def special_tokens_map_extended(self) -> Dict[str, Union[str, AddedToken, List[Union[str, AddedToken]]]]:
        """
        `Dict[str, Union[str, tokenizers.AddedToken, List[Union[str, tokenizers.AddedToken]]]]`: A dictionary mapping
        special token class attributes (`cls_token`, `unk_token`, etc.) to their values (`'<unk>'`, `'<cls>'`, etc.).

        Don't convert tokens of `tokenizers.AddedToken` type to string so they can be used to control more finely how
        special tokens are tokenized.
        """
        set_attr = {}
        for attr in self.SPECIAL_TOKENS_ATTRIBUTES:
            attr_value = getattr(self, "_" + attr)
            if attr_value:
                set_attr[attr] = attr_value
        return set_attr

    @property
    def all_special_tokens_extended(self) -> List[Union[str, AddedToken]]:
        """
        `List[Union[str, tokenizers.AddedToken]]`: All the special tokens (`'<unk>'`, `'<cls>'`, etc.), the order has
        nothing to do with the index of each tokens. If you want to know the correct indices, check
        `self.added_tokens_encoder`. We can't create an order anymore as the keys are `AddedTokens` and not `Strings`.

        Don't convert tokens of `tokenizers.AddedToken` type to string so they can be used to control more finely how
        special tokens are tokenized.
        """
        all_tokens = []
        seen = set()
        for value in self.special_tokens_map_extended.values():
            if isinstance(value, (list, tuple)):
                tokens_to_add = [token for token in value if str(token) not in seen]
            else:
                tokens_to_add = [value] if str(value) not in seen else []
            seen.update(map(str, tokens_to_add))
            all_tokens.extend(tokens_to_add)
        return all_tokens

    @property
    def all_special_tokens(self) -> List[str]:
        """
        `List[str]`: A list of the unique special tokens (`'<unk>'`, `'<cls>'`, ..., etc.).

        Convert tokens of `tokenizers.AddedToken` type to string.
        """
        all_toks = [str(s) for s in self.all_special_tokens_extended]
        return all_toks

    @property
    def all_special_ids(self) -> List[int]:
        """
        `List[int]`: List the ids of the special tokens(`'<unk>'`, `'<cls>'`, etc.) mapped to class attributes.
        """
        all_toks = self.all_special_tokens
        all_ids = self.convert_tokens_to_ids(all_toks)
        return all_ids

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.additional_special_tokens: List[str] property writable

List[str]: All the additional special tokens you may want to use. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.additional_special_tokens_ids: List[int] property writable

List[int]: Ids of all the additional special tokens in the vocabulary. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.all_special_ids: List[int] property

List[int]: List the ids of the special tokens('<unk>', '<cls>', etc.) mapped to class attributes.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.all_special_tokens: List[str] property

List[str]: A list of the unique special tokens ('<unk>', '<cls>', ..., etc.).

Convert tokens of tokenizers.AddedToken type to string.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.all_special_tokens_extended: List[Union[str, AddedToken]] property

List[Union[str, tokenizers.AddedToken]]: All the special tokens ('<unk>', '<cls>', etc.), the order has nothing to do with the index of each tokens. If you want to know the correct indices, check self.added_tokens_encoder. We can't create an order anymore as the keys are AddedTokens and not Strings.

Don't convert tokens of tokenizers.AddedToken type to string so they can be used to control more finely how special tokens are tokenized.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.bos_token: str property writable

str: Beginning of sentence token. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.bos_token_id: Optional[int] property writable

Optional[int]: Id of the beginning of sentence token in the vocabulary. Returns None if the token has not been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.cls_token: str property writable

str: Classification token, to extract a summary of an input sequence leveraging self-attention along the full depth of the model. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.cls_token_id: Optional[int] property writable

Optional[int]: Id of the classification token in the vocabulary, to extract a summary of an input sequence leveraging self-attention along the full depth of the model.

Returns None if the token has not been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.eos_token: str property writable

str: End of sentence token. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.eos_token_id: Optional[int] property writable

Optional[int]: Id of the end of sentence token in the vocabulary. Returns None if the token has not been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.mask_token: str property writable

str: Mask token, to use when training a model with masked-language modeling. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.mask_token_id: Optional[int] property writable

Optional[int]: Id of the mask token in the vocabulary, used when training a model with masked-language modeling. Returns None if the token has not been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.pad_token: str property writable

str: Padding token. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.pad_token_id: Optional[int] property writable

Optional[int]: Id of the padding token in the vocabulary. Returns None if the token has not been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.pad_token_type_id: int property

int: Id of the padding token type in the vocabulary.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.sep_token: str property writable

str: Separation token, to separate context and query in an input sequence. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.sep_token_id: Optional[int] property writable

Optional[int]: Id of the separation token in the vocabulary, to separate context and query in an input sequence. Returns None if the token has not been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.special_tokens_map: Dict[str, Union[str, List[str]]] property

Dict[str, Union[str, List[str]]]: A dictionary mapping special token class attributes (cls_token, unk_token, etc.) to their values ('<unk>', '<cls>', etc.).

Convert potential tokens of tokenizers.AddedToken type to string.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.special_tokens_map_extended: Dict[str, Union[str, AddedToken, List[Union[str, AddedToken]]]] property

Dict[str, Union[str, tokenizers.AddedToken, List[Union[str, tokenizers.AddedToken]]]]: A dictionary mapping special token class attributes (cls_token, unk_token, etc.) to their values ('<unk>', '<cls>', etc.).

Don't convert tokens of tokenizers.AddedToken type to string so they can be used to control more finely how special tokens are tokenized.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.unk_token: str property writable

str: Unknown token. Log an error if used while not having been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.unk_token_id: Optional[int] property writable

Optional[int]: Id of the unknown token in the vocabulary. Returns None if the token has not been set.

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.add_special_tokens(special_tokens_dict, replace_additional_special_tokens=True)

Add a dictionary of special tokens (eos, pad, cls, etc.) to the encoder and link them to class attributes. If special tokens are NOT in the vocabulary, they are added to it (indexed starting from the last index of the current vocabulary).

When adding new tokens to the vocabulary, you should make sure to also resize the token embedding matrix of the model so that its embedding matrix matches the tokenizer.

In order to do that, please use the [~PreTrainedModel.resize_token_embeddings] method.

Using add_special_tokens will ensure your special tokens can be used in several ways:

• Special tokens can be skipped when decoding using skip_special_tokens = True.
• Special tokens are carefully handled by the tokenizer (they are never split), similar to AddedTokens.
• You can easily refer to special tokens using tokenizer class attributes like tokenizer.cls_token. This makes it easy to develop model-agnostic training and fine-tuning scripts.

When possible, special tokens are already registered for provided pretrained models (for instance [BertTokenizer] cls_token is already registered to be :obj*'[CLS]'* and XLM's one is also registered to be '</s>').

PARAMETER	DESCRIPTION
special_tokens_dict	Keys should be in the list of predefined special attributes: [bos_token, eos_token, unk_token, sep_token, pad_token, cls_token, mask_token, additional_special_tokens]. Tokens are only added if they are not already in the vocabulary (tested by checking if the tokenizer assign the index of the unk_token to them). TYPE: dictionary *str* to *str* or `tokenizers.AddedToken`
replace_additional_special_tokens	If True, the existing list of additional special tokens will be replaced by the list provided in special_tokens_dict. Otherwise, self._additional_special_tokens is just extended. In the former case, the tokens will NOT be removed from the tokenizer's full vocabulary - they are only being flagged as non-special tokens. Remember, this only affects which tokens are skipped during decoding, not the added_tokens_encoder and added_tokens_decoder. This means that the previous additional_special_tokens are still added tokens, and will not be split by the model. TYPE: `bool`, *optional*,, defaults to `True` DEFAULT: True

RETURNS	DESCRIPTION
int	int: Number of tokens added to the vocabulary.

# Let's see how to add a new classification token to GPT-2
tokenizer = GPT2Tokenizer.from_pretrained("openai-community/gpt2")
model = GPT2Model.from_pretrained("openai-community/gpt2")

special_tokens_dict = {"cls_token": "<CLS>"}

num_added_toks = tokenizer.add_special_tokens(special_tokens_dict)
print("We have added", num_added_toks, "tokens")
# Notice: resize_token_embeddings expect to receive the full size of the new vocabulary, i.e., the length of the tokenizer.
model.resize_token_embeddings(len(tokenizer))

assert tokenizer.cls_token == "<CLS>"

Source code in mindnlp\transformers\tokenization_utils_base.py

def add_special_tokens(
    self, special_tokens_dict: Dict[str, Union[str, AddedToken]], replace_additional_special_tokens=True
) -> int:
    """
    Add a dictionary of special tokens (eos, pad, cls, etc.) to the encoder and link them to class attributes. If
    special tokens are NOT in the vocabulary, they are added to it (indexed starting from the last index of the
    current vocabulary).

    When adding new tokens to the vocabulary, you should make sure to also resize the token embedding matrix of the
    model so that its embedding matrix matches the tokenizer.

    In order to do that, please use the [`~PreTrainedModel.resize_token_embeddings`] method.

    Using `add_special_tokens` will ensure your special tokens can be used in several ways:

    - Special tokens can be skipped when decoding using `skip_special_tokens = True`.
    - Special tokens are carefully handled by the tokenizer (they are never split), similar to `AddedTokens`.
    - You can easily refer to special tokens using tokenizer class attributes like `tokenizer.cls_token`. This
      makes it easy to develop model-agnostic training and fine-tuning scripts.

    When possible, special tokens are already registered for provided pretrained models (for instance
    [`BertTokenizer`] `cls_token` is already registered to be :obj*'[CLS]'* and XLM's one is also registered to be
    `'</s>'`).

    Args:
        special_tokens_dict (dictionary *str* to *str* or `tokenizers.AddedToken`):
            Keys should be in the list of predefined special attributes: [`bos_token`, `eos_token`, `unk_token`,
            `sep_token`, `pad_token`, `cls_token`, `mask_token`, `additional_special_tokens`].

            Tokens are only added if they are not already in the vocabulary (tested by checking if the tokenizer
            assign the index of the `unk_token` to them).
        replace_additional_special_tokens (`bool`, *optional*,, defaults to `True`):
            If `True`, the existing list of additional special tokens will be replaced by the list provided in
            `special_tokens_dict`. Otherwise, `self._additional_special_tokens` is just extended. In the former
            case, the tokens will NOT be removed from the tokenizer's full vocabulary - they are only being flagged
            as non-special tokens. Remember, this only affects which tokens are skipped during decoding, not the
            `added_tokens_encoder` and `added_tokens_decoder`. This means that the previous
            `additional_special_tokens` are still added tokens, and will not be split by the model.

    Returns:
        `int`: Number of tokens added to the vocabulary.

    Examples:

    ```python
    # Let's see how to add a new classification token to GPT-2
    tokenizer = GPT2Tokenizer.from_pretrained("openai-community/gpt2")
    model = GPT2Model.from_pretrained("openai-community/gpt2")

    special_tokens_dict = {"cls_token": "<CLS>"}

    num_added_toks = tokenizer.add_special_tokens(special_tokens_dict)
    print("We have added", num_added_toks, "tokens")
    # Notice: resize_token_embeddings expect to receive the full size of the new vocabulary, i.e., the length of the tokenizer.
    model.resize_token_embeddings(len(tokenizer))

    assert tokenizer.cls_token == "<CLS>"
    ```"""
    if not special_tokens_dict:
        return 0

    added_tokens = []
    for key, value in special_tokens_dict.items():
        assert key in self.SPECIAL_TOKENS_ATTRIBUTES, f"Key {key} is not a special token"

        if self.verbose:
            logger.info(f"Assigning {value} to the {key} key of the tokenizer")

        if key == "additional_special_tokens":
            assert isinstance(value, (list, tuple)) and all(
                isinstance(t, (str, AddedToken)) for t in value
            ), f"Tokens {value} for key {key} should all be str or AddedToken instances"

            to_add = []
            for token in value:
                if isinstance(token, str):
                    # for legacy purpose we default to stripping. `test_add_tokens_tokenizer` depends on this
                    token = AddedToken(token, rstrip=False, lstrip=False, normalized=False, special=True)
                if not replace_additional_special_tokens and str(token) in self.additional_special_tokens:
                    continue
                to_add.append(token)
            if replace_additional_special_tokens and len(to_add) > 0:
                setattr(self, key, list(to_add))
            else:
                self._additional_special_tokens.extend(to_add)
            added_tokens += to_add

        else:
            if not isinstance(value, (str, AddedToken)):
                raise ValueError(f"Token {value} for key {key} should be a str or an AddedToken instance")
            if isinstance(value, (str)):
                # for legacy purpose we default to stripping. `False` depends on this
                value = AddedToken(value, rstrip=False, lstrip=False, normalized=False, special=True)
            if isinstance(value, AddedToken):
                setattr(self, key, value)
            if value not in added_tokens:
                added_tokens.append(value)

    # if we are adding tokens that were not part of the vocab, we ought to add them
    added_tokens = self.add_tokens(added_tokens, special_tokens=True)
    return added_tokens

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.add_tokens(new_tokens, special_tokens=False)

Add a list of new tokens to the tokenizer class. If the new tokens are not in the vocabulary, they are added to it with indices starting from length of the current vocabulary and and will be isolated before the tokenization algorithm is applied. Added tokens and tokens from the vocabulary of the tokenization algorithm are therefore not treated in the same way.

Note, when adding new tokens to the vocabulary, you should make sure to also resize the token embedding matrix of the model so that its embedding matrix matches the tokenizer.

In order to do that, please use the [~PreTrainedModel.resize_token_embeddings] method.

PARAMETER	DESCRIPTION
new_tokens	Tokens are only added if they are not already in the vocabulary. tokenizers.AddedToken wraps a string token to let you personalize its behavior: whether this token should only match against a single word, whether this token should strip all potential whitespaces on the left side, whether this token should strip all potential whitespaces on the right side, etc. TYPE: `str`, `tokenizers.AddedToken` or a list of *str* or `tokenizers.AddedToken`
special_tokens	Can be used to specify if the token is a special token. This mostly change the normalization behavior (special tokens like CLS or [MASK] are usually not lower-cased for instance). See details for tokenizers.AddedToken in HuggingFace tokenizers library. TYPE: `bool`, *optional*, defaults to `False` DEFAULT: False

RETURNS	DESCRIPTION
int	int: Number of tokens added to the vocabulary.

# Let's see how to increase the vocabulary of Bert model and tokenizer
tokenizer = BertTokenizerFast.from_pretrained("google-bert/bert-base-uncased")
model = BertModel.from_pretrained("google-bert/bert-base-uncased")

num_added_toks = tokenizer.add_tokens(["new_tok1", "my_new-tok2"])
print("We have added", num_added_toks, "tokens")
# Notice: resize_token_embeddings expect to receive the full size of the new vocabulary, i.e., the length of the tokenizer.
model.resize_token_embeddings(len(tokenizer))

Source code in mindnlp\transformers\tokenization_utils_base.py

def add_tokens(
    self, new_tokens: Union[str, AddedToken, List[Union[str, AddedToken]]], special_tokens: bool = False
) -> int:
    """
    Add a list of new tokens to the tokenizer class. If the new tokens are not in the vocabulary, they are added to
    it with indices starting from length of the current vocabulary and and will be isolated before the tokenization
    algorithm is applied. Added tokens and tokens from the vocabulary of the tokenization algorithm are therefore
    not treated in the same way.

    Note, when adding new tokens to the vocabulary, you should make sure to also resize the token embedding matrix
    of the model so that its embedding matrix matches the tokenizer.

    In order to do that, please use the [`~PreTrainedModel.resize_token_embeddings`] method.

    Args:
        new_tokens (`str`, `tokenizers.AddedToken` or a list of *str* or `tokenizers.AddedToken`):
            Tokens are only added if they are not already in the vocabulary. `tokenizers.AddedToken` wraps a string
            token to let you personalize its behavior: whether this token should only match against a single word,
            whether this token should strip all potential whitespaces on the left side, whether this token should
            strip all potential whitespaces on the right side, etc.
        special_tokens (`bool`, *optional*, defaults to `False`):
            Can be used to specify if the token is a special token. This mostly change the normalization behavior
            (special tokens like CLS or [MASK] are usually not lower-cased for instance).

            See details for `tokenizers.AddedToken` in HuggingFace tokenizers library.

    Returns:
        `int`: Number of tokens added to the vocabulary.

    Examples:

    ```python
    # Let's see how to increase the vocabulary of Bert model and tokenizer
    tokenizer = BertTokenizerFast.from_pretrained("google-bert/bert-base-uncased")
    model = BertModel.from_pretrained("google-bert/bert-base-uncased")

    num_added_toks = tokenizer.add_tokens(["new_tok1", "my_new-tok2"])
    print("We have added", num_added_toks, "tokens")
    # Notice: resize_token_embeddings expect to receive the full size of the new vocabulary, i.e., the length of the tokenizer.
    model.resize_token_embeddings(len(tokenizer))
    ```"""
    if not new_tokens:
        return 0

    if not isinstance(new_tokens, (list, tuple)):
        new_tokens = [new_tokens]

    return self._add_tokens(new_tokens, special_tokens=special_tokens)

mindnlp.transformers.tokenization_utils_base.SpecialTokensMixin.sanitize_special_tokens()

The sanitize_special_tokens is now deprecated kept for backward compatibility and will be removed in transformers v5.

Source code in mindnlp\transformers\tokenization_utils_base.py

def sanitize_special_tokens(self) -> int:
    """
    The `sanitize_special_tokens` is now deprecated kept for backward compatibility and will be removed in
    transformers v5.
    """
    logger.warning_once("The `sanitize_special_tokens` will be removed in transformers v5.")
    return self.add_tokens(self.all_special_tokens_extended, special_tokens=True)