medcat.components.addons.relation_extraction.llama.tokenizer

Attributes

logger

Classes

ConfigRelCAT	The RelCAT part of the config
BaseTokenizerWrapper	Base class for all fast tokenizers (wrapping HuggingFace tokenizers library).
TokenizerWrapperLlama	Wrapper around a huggingface Llama tokenizer so that it works with the

Module Contents

class medcat.components.addons.relation_extraction.llama.tokenizer.ConfigRelCAT(/, **data)

Bases: medcat.config.config.ComponentConfig

The RelCAT part of the config

Parameters:

data (Any)

general: General

model: Model

train: Train

class Config

extra = 'allow'

validate_assignment = True

classmethod load(load_path='./')

Load the config from a file.

Parameters:

load_path (str) – Path to RelCAT config. Defaults to “./”.

Returns:

ConfigRelCAT – The loaded config.

Return type:

ConfigRelCAT

comp_name: str = 'default'

The name of the component.

If a custom implementation is required, it needs to be registered using `medcat.components.types.register_core_component(

<core component type>, <component name>, <implementing class>)

By default, only the ‘default’ component is registered.

_is_dirty: bool = False

__setattr__(name, value)

Parameters:

• name (str)

• value (Any)

property is_dirty: bool

Return type:

bool

mark_clean()

get_strategy()

Return type:

medcat.storage.serialisables.SerialisingStrategy

classmethod get_init_attrs()

Return type:

list[str]

classmethod ignore_attrs()

Return type:

list[str]

classmethod include_properties()

Return type:

list[str]

merge_config(other)

Merge this config with another config’s (partial) model dump.

The exepctation is that the other dict is a partial model dump. Values specified there are overwritten into the current config. Values not specified there are left intact.

The other config can have keys/values that do not exist in the config or sub-config. And they will be added where possible.

Parameters:

other (dict) – The model dump

Raises:

IncorrectConfigValues – If unable to set the attribute, trying to set incorrect value, or trying to set sub-config values in an incorrect format (non-dict).

model_config: ClassVar[pydantic.config.ConfigDict]

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

model_fields: ClassVar[Dict[str, pydantic.fields.FieldInfo]]

Metadata about the fields defined on the model, mapping of field names to [FieldInfo][pydantic.fields.FieldInfo] objects.

This replaces Model.__fields__ from Pydantic V1.

model_computed_fields: ClassVar[Dict[str, pydantic.fields.ComputedFieldInfo]]

A dictionary of computed field names and their corresponding ComputedFieldInfo objects.

__class_vars__: ClassVar[set[str]]

The names of the class variables defined on the model.

__private_attributes__: ClassVar[Dict[str, pydantic.fields.ModelPrivateAttr]]

Metadata about the private attributes of the model.

__signature__: ClassVar[inspect.Signature]

The synthesized __init__ [Signature][inspect.Signature] of the model.

__pydantic_complete__: ClassVar[bool] = False

Whether model building is completed, or if there are still undefined fields.

__pydantic_core_schema__: ClassVar[pydantic_core.CoreSchema]

The core schema of the model.

__pydantic_custom_init__: ClassVar[bool]

Whether the model has a custom __init__ method.

__pydantic_decorators__: ClassVar[pydantic._internal._decorators.DecoratorInfos]

Metadata containing the decorators defined on the model. This replaces Model.__validators__ and Model.__root_validators__ from Pydantic V1.

__pydantic_generic_metadata__: ClassVar[pydantic._internal._generics.PydanticGenericMetadata]

Metadata for generic models; contains data used for a similar purpose to __args__, __origin__, __parameters__ in typing-module generics. May eventually be replaced by these.

__pydantic_parent_namespace__: ClassVar[Dict[str, Any] | None] = None

Parent namespace of the model, used for automatic rebuilding of models.

__pydantic_post_init__: ClassVar[None | Literal['model_post_init']]

The name of the post-init method for the model, if defined.

__pydantic_root_model__: ClassVar[bool] = False

Whether the model is a [RootModel][pydantic.root_model.RootModel].

__pydantic_serializer__: ClassVar[pydantic_core.SchemaSerializer]

The pydantic-core SchemaSerializer used to dump instances of the model.

__pydantic_validator__: ClassVar[pydantic_core.SchemaValidator | pydantic.plugin._schema_validator.PluggableSchemaValidator]

The pydantic-core SchemaValidator used to validate instances of the model.

__pydantic_extra__: dict[str, Any] | None

A dictionary containing extra values, if [extra][pydantic.config.ConfigDict.extra] is set to ‘allow’.

__pydantic_fields_set__: set[str]

The names of fields explicitly set during instantiation.

__pydantic_private__: dict[str, Any] | None

Values of private attributes set on the model instance.

__slots__ = ('__dict__', '__pydantic_fields_set__', '__pydantic_extra__', '__pydantic_private__')

__init__(/, **data)

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:

data (Any)

Return type:

None

property model_extra: dict[str, Any] | None

Get extra fields set during validation.

Returns:

A dictionary of extra fields, or `None` if `config.extra` is not set to `”allow”`.

Return type:

dict[str, Any] | None

property model_fields_set: set[str]

Returns the set of fields that have been explicitly set on this model instance.

Returns:

A set of strings representing the fields that have been set, – i.e. that were not filled from defaults.

Return type:

set[str]

classmethod model_construct(_fields_set=None, **values)

Creates a new instance of the Model class with validated data.

Creates a new model setting __dict__ and __pydantic_fields_set__ from trusted or pre-validated data. Default values are respected, but no other validation is performed.

!!! note

model_construct() generally respects the model_config.extra setting on the provided model. That is, if model_config.extra == ‘allow’, then all extra passed values are added to the model instance’s __dict__ and __pydantic_extra__ fields. If model_config.extra == ‘ignore’ (the default), then all extra passed values are ignored. Because no validation is performed with a call to model_construct(), having model_config.extra == ‘forbid’ does not result in an error if extra values are passed, but they will be ignored.

Parameters:

• _fields_set (set[str] | None) – A set of field names that were originally explicitly set during instantiation. If provided, this is directly used for the [model_fields_set][pydantic.BaseModel.model_fields_set] attribute. Otherwise, the field names from the values argument will be used.

• values (Any) – Trusted or pre-validated data dictionary.

Returns:

A new instance of the `Model` class with validated data.

Return type:

typing_extensions.Self

model_copy(*, update=None, deep=False)

Usage docs: https://docs.pydantic.dev/2.9/concepts/serialization/#model_copy

Returns a copy of the model.

Parameters:

• update (dict[str, Any] | None) – Values to change/add in the new model. Note: the data is not validated before creating the new model. You should trust this data.

• deep (bool) – Set to True to make a deep copy of the model.

Returns:

New model instance.

Return type:

typing_extensions.Self

model_dump(*, mode='python', include=None, exclude=None, context=None, by_alias=False, exclude_unset=False, exclude_defaults=False, exclude_none=False, round_trip=False, warnings=True, serialize_as_any=False)

Usage docs: https://docs.pydantic.dev/2.9/concepts/serialization/#modelmodel_dump

Generate a dictionary representation of the model, optionally specifying which fields to include or exclude.

Parameters:

• mode (Literal['json', 'python'] | str) – The mode in which to_python should run. If mode is ‘json’, the output will only contain JSON serializable types. If mode is ‘python’, the output may contain non-JSON-serializable Python objects.

• include (IncEx | None) – A set of fields to include in the output.

• exclude (IncEx | None) – A set of fields to exclude from the output.

• context (Any | None) – Additional context to pass to the serializer.

• by_alias (bool) – Whether to use the field’s alias in the dictionary key if defined.

• exclude_unset (bool) – Whether to exclude fields that have not been explicitly set.

• exclude_defaults (bool) – Whether to exclude fields that are set to their default value.

• exclude_none (bool) – Whether to exclude fields that have a value of None.

• round_trip (bool) – If True, dumped values should be valid as input for non-idempotent types such as Json[T].

• warnings (bool | Literal['none', 'warn', 'error']) – How to handle serialization errors. False/”none” ignores them, True/”warn” logs errors, “error” raises a [PydanticSerializationError][pydantic_core.PydanticSerializationError].

• serialize_as_any (bool) – Whether to serialize fields with duck-typing serialization behavior.

Returns:

A dictionary representation of the model.

Return type:

dict[str, Any]

model_dump_json(*, indent=None, include=None, exclude=None, context=None, by_alias=False, exclude_unset=False, exclude_defaults=False, exclude_none=False, round_trip=False, warnings=True, serialize_as_any=False)

Usage docs: https://docs.pydantic.dev/2.9/concepts/serialization/#modelmodel_dump_json

Generates a JSON representation of the model using Pydantic’s to_json method.

Parameters:

• indent (int | None) – Indentation to use in the JSON output. If None is passed, the output will be compact.

• include (IncEx | None) – Field(s) to include in the JSON output.

• exclude (IncEx | None) – Field(s) to exclude from the JSON output.

• context (Any | None) – Additional context to pass to the serializer.

• by_alias (bool) – Whether to serialize using field aliases.

• exclude_unset (bool) – Whether to exclude fields that have not been explicitly set.

• exclude_defaults (bool) – Whether to exclude fields that are set to their default value.

• exclude_none (bool) – Whether to exclude fields that have a value of None.

• round_trip (bool) – If True, dumped values should be valid as input for non-idempotent types such as Json[T].

• warnings (bool | Literal['none', 'warn', 'error']) – How to handle serialization errors. False/”none” ignores them, True/”warn” logs errors, “error” raises a [PydanticSerializationError][pydantic_core.PydanticSerializationError].

• serialize_as_any (bool) – Whether to serialize fields with duck-typing serialization behavior.

Returns:

A JSON string representation of the model.

Return type:

str

classmethod model_json_schema(by_alias=True, ref_template=DEFAULT_REF_TEMPLATE, schema_generator=GenerateJsonSchema, mode='validation')

Generates a JSON schema for a model class.

Parameters:

• by_alias (bool) – Whether to use attribute aliases or not.

• ref_template (str) – The reference template.

• schema_generator (type[pydantic.json_schema.GenerateJsonSchema]) – To override the logic used to generate the JSON schema, as a subclass of GenerateJsonSchema with your desired modifications

• mode (pydantic.json_schema.JsonSchemaMode) – The mode in which to generate the schema.

Returns:

The JSON schema for the given model class.

Return type:

dict[str, Any]

classmethod model_parametrized_name(params)

Compute the class name for parametrizations of generic classes.

This method can be overridden to achieve a custom naming scheme for generic BaseModels.

Parameters:

params (tuple[type[Any], Ellipsis]) – Tuple of types of the class. Given a generic class Model with 2 type variables and a concrete model Model[str, int], the value (str, int) would be passed to params.

Returns:

String representing the new class where `params` are passed to `cls` as type variables.

Raises:

TypeError – Raised when trying to generate concrete names for non-generic models.

Return type:

str

model_post_init(__context)

Override this method to perform additional initialization after __init__ and model_construct. This is useful if you want to do some validation that requires the entire model to be initialized.

Parameters:

__context (Any)

Return type:

None

classmethod model_rebuild(*, force=False, raise_errors=True, _parent_namespace_depth=2, _types_namespace=None)

Try to rebuild the pydantic-core schema for the model.

This may be necessary when one of the annotations is a ForwardRef which could not be resolved during the initial attempt to build the schema, and automatic rebuilding fails.

Parameters:

• force (bool) – Whether to force the rebuilding of the model schema, defaults to False.

• raise_errors (bool) – Whether to raise errors, defaults to True.

• _parent_namespace_depth (int) – The depth level of the parent namespace, defaults to 2.

• _types_namespace (dict[str, Any] | None) – The types namespace, defaults to None.

Returns:

• Returns `None` if the schema is already “complete” and rebuilding was not required.

• If rebuilding _was_ required, returns `True` if rebuilding was successful, otherwise `False`.

Return type:

bool | None

classmethod model_validate(obj, *, strict=None, from_attributes=None, context=None)

Validate a pydantic model instance.

Parameters:

• obj (Any) – The object to validate.

• strict (bool | None) – Whether to enforce types strictly.

• from_attributes (bool | None) – Whether to extract data from object attributes.

• context (Any | None) – Additional context to pass to the validator.

Raises:

ValidationError – If the object could not be validated.

Returns:

The validated model instance.

Return type:

typing_extensions.Self

classmethod model_validate_json(json_data, *, strict=None, context=None)

Usage docs: https://docs.pydantic.dev/2.9/concepts/json/#json-parsing

Validate the given JSON data against the Pydantic model.

Parameters:

• json_data (str | bytes | bytearray) – The JSON data to validate.

• strict (bool | None) – Whether to enforce types strictly.

• context (Any | None) – Extra variables to pass to the validator.

Returns:

The validated Pydantic model.

Raises:

ValidationError – If json_data is not a JSON string or the object could not be validated.

Return type:

typing_extensions.Self

classmethod model_validate_strings(obj, *, strict=None, context=None)

Validate the given object with string data against the Pydantic model.

Parameters:

• obj (Any) – The object containing string data to validate.

• strict (bool | None) – Whether to enforce types strictly.

• context (Any | None) – Extra variables to pass to the validator.

Returns:

The validated Pydantic model.

Return type:

typing_extensions.Self

classmethod __get_pydantic_core_schema__(source, handler, /)

Hook into generating the model’s CoreSchema.

Parameters:

• source (type[BaseModel]) – The class we are generating a schema for. This will generally be the same as the cls argument if this is a classmethod.

• handler (pydantic.annotated_handlers.GetCoreSchemaHandler) – A callable that calls into Pydantic’s internal CoreSchema generation logic.

Returns:

A `pydantic-core` `CoreSchema`.

Return type:

pydantic_core.CoreSchema

classmethod __get_pydantic_json_schema__(core_schema, handler, /)

Hook into generating the model’s JSON schema.

Parameters:

• core_schema (pydantic_core.CoreSchema) – A pydantic-core CoreSchema. You can ignore this argument and call the handler with a new CoreSchema, wrap this CoreSchema ({‘type’: ‘nullable’, ‘schema’: current_schema}), or just call the handler with the original schema.

• handler (pydantic.annotated_handlers.GetJsonSchemaHandler) – Call into Pydantic’s internal JSON schema generation. This will raise a pydantic.errors.PydanticInvalidForJsonSchema if JSON schema generation fails. Since this gets called by BaseModel.model_json_schema you can override the schema_generator argument to that function to change JSON schema generation globally for a type.

Returns:

A JSON schema, as a Python object.

Return type:

pydantic.json_schema.JsonSchemaValue

classmethod __pydantic_init_subclass__(**kwargs)

This is intended to behave just like __init_subclass__, but is called by ModelMetaclass only after the class is actually fully initialized. In particular, attributes like model_fields will be present when this is called.

This is necessary because __init_subclass__ will always be called by type.__new__, and it would require a prohibitively large refactor to the ModelMetaclass to ensure that type.__new__ was called in such a manner that the class would already be sufficiently initialized.

This will receive the same kwargs that would be passed to the standard __init_subclass__, namely, any kwargs passed to the class definition that aren’t used internally by pydantic.

Parameters:

**kwargs (Any) – Any keyword arguments passed to the class definition that aren’t used internally by pydantic.

Return type:

None

classmethod __class_getitem__(typevar_values)

Parameters:

typevar_values (type[Any] | tuple[type[Any], Ellipsis])

Return type:

type[BaseModel] | pydantic._internal._forward_ref.PydanticRecursiveRef

__copy__()

Returns a shallow copy of the model.

Return type:

typing_extensions.Self

__deepcopy__(memo=None)

Returns a deep copy of the model.

Parameters:

memo (dict[int, Any] | None)

Return type:

typing_extensions.Self

__getattr__(item)

Parameters:

item (str)

Return type:

Any

_check_frozen(name, value)

Parameters:

• name (str)

• value (Any)

Return type:

None

__getstate__()

Return type:

dict[Any, Any]

__setstate__(state)

Parameters:

state (dict[Any, Any])

Return type:

None

__eq__(other)

Parameters:

other (Any)

Return type:

bool

classmethod __init_subclass__(**kwargs)

This signature is included purely to help type-checkers check arguments to class declaration, which provides a way to conveniently set model_config key/value pairs.

```py from pydantic import BaseModel

class MyModel(BaseModel, extra=’allow’): … ```

However, this may be deceiving, since the _actual_ calls to __init_subclass__ will not receive any of the config arguments, and will only receive any keyword arguments passed during class initialization that are _not_ expected keys in ConfigDict. (This is due to the way ModelMetaclass.__new__ works.)

Parameters:

**kwargs (typing_extensions.Unpack[pydantic.config.ConfigDict]) – Keyword arguments passed to the class definition, which set model_config

Note

You may want to override __pydantic_init_subclass__ instead, which behaves similarly but is called after the class is fully initialized.

__iter__()

So dict(model) works.

Return type:

TupleGenerator

__repr__()

Return type:

str

__repr_args__()

Return type:

pydantic._internal._repr.ReprArgs

__repr_name__

__repr_str__

__pretty__

__rich_repr__

__str__()

Return type:

str

property __fields__: dict[str, pydantic.fields.FieldInfo]

Return type:

dict[str, pydantic.fields.FieldInfo]

property __fields_set__: set[str]

Return type:

set[str]

dict(*, include=None, exclude=None, by_alias=False, exclude_unset=False, exclude_defaults=False, exclude_none=False)

Parameters:

• include (IncEx | None)

• exclude (IncEx | None)

• by_alias (bool)

• exclude_unset (bool)

• exclude_defaults (bool)

• exclude_none (bool)

Return type:

Dict[str, Any]

json(*, include=None, exclude=None, by_alias=False, exclude_unset=False, exclude_defaults=False, exclude_none=False, encoder=PydanticUndefined, models_as_dict=PydanticUndefined, **dumps_kwargs)

Parameters:

• include (IncEx | None)

• exclude (IncEx | None)

• by_alias (bool)

• exclude_unset (bool)

• exclude_defaults (bool)

• exclude_none (bool)

• encoder (Callable[[Any], Any] | None)

• models_as_dict (bool)

• dumps_kwargs (Any)

Return type:

str

classmethod parse_obj(obj)

Parameters:

obj (Any)

Return type:

typing_extensions.Self

classmethod parse_raw(b, *, content_type=None, encoding='utf8', proto=None, allow_pickle=False)

Parameters:

• b (str | bytes)

• content_type (str | None)

• encoding (str)

• proto (pydantic.deprecated.parse.Protocol | None)

• allow_pickle (bool)

Return type:

typing_extensions.Self

classmethod parse_file(path, *, content_type=None, encoding='utf8', proto=None, allow_pickle=False)

Parameters:

• path (str | pathlib.Path)

• content_type (str | None)

• encoding (str)

• proto (pydantic.deprecated.parse.Protocol | None)

• allow_pickle (bool)

Return type:

typing_extensions.Self

classmethod from_orm(obj)

Parameters:

obj (Any)

Return type:

typing_extensions.Self

classmethod construct(_fields_set=None, **values)

Parameters:

• _fields_set (set[str] | None)

• values (Any)

Return type:

typing_extensions.Self

copy(*, include=None, exclude=None, update=None, deep=False)

Returns a copy of the model.

!!! warning “Deprecated”

This method is now deprecated; use model_copy instead.

If you need include or exclude, use:

`py data = self.model_dump(include=include, exclude=exclude, round_trip=True) data = {**data, **(update or {})} copied = self.model_validate(data) `

Parameters:

• include (pydantic._internal._utils.AbstractSetIntStr | pydantic._internal._utils.MappingIntStrAny | None) – Optional set or mapping specifying which fields to include in the copied model.

• exclude (pydantic._internal._utils.AbstractSetIntStr | pydantic._internal._utils.MappingIntStrAny | None) – Optional set or mapping specifying which fields to exclude in the copied model.

• update (Dict[str, Any] | None) – Optional dictionary of field-value pairs to override field values in the copied model.

• deep (bool) – If True, the values of fields that are Pydantic models will be deep-copied.

Returns:

A copy of the model with included, excluded and updated fields as specified.

Return type:

typing_extensions.Self

classmethod schema(by_alias=True, ref_template=DEFAULT_REF_TEMPLATE)

Parameters:

• by_alias (bool)

• ref_template (str)

Return type:

Dict[str, Any]

classmethod schema_json(*, by_alias=True, ref_template=DEFAULT_REF_TEMPLATE, **dumps_kwargs)

Parameters:

• by_alias (bool)

• ref_template (str)

• dumps_kwargs (Any)

Return type:

str

classmethod validate(value)

Parameters:

value (Any)

Return type:

typing_extensions.Self

classmethod update_forward_refs(**localns)

Parameters:

localns (Any)

Return type:

None

_iter(*args, **kwargs)

Parameters:

• args (Any)

• kwargs (Any)

Return type:

Any

_copy_and_set_values(*args, **kwargs)

Parameters:

• args (Any)

• kwargs (Any)

Return type:

Any

classmethod _get_value(*args, **kwargs)

Parameters:

• args (Any)

• kwargs (Any)

Return type:

Any

_calculate_keys(*args, **kwargs)

Parameters:

• args (Any)

• kwargs (Any)

Return type:

Any

class medcat.components.addons.relation_extraction.llama.tokenizer.BaseTokenizerWrapper(hf_tokenizers=None, max_seq_length=None, add_special_tokens=False)

Bases: transformers.PreTrainedTokenizerFast

Base class for all fast tokenizers (wrapping HuggingFace tokenizers library).

Inherits from [~tokenization_utils_base.PreTrainedTokenizerBase].

Handles all the shared methods for tokenization and special tokens, as well as methods for downloading/caching/loading pretrained tokenizers, as well as adding tokens to the vocabulary.

This class also contains the added tokens in a unified way on top of all tokenizers so we don’t have to handle the specific vocabulary augmentation methods of the various underlying dictionary structures (BPE, sentencepiece…).

Parameters:

• max_seq_length (Optional[int])

• add_special_tokens (Optional[bool])

name = 'base_tokenizer_wrapper_rel'

__init__(hf_tokenizers=None, max_seq_length=None, add_special_tokens=False)

Parameters:

• max_seq_length (Optional[int])

• add_special_tokens (Optional[bool])

hf_tokenizers = None

max_seq_length = None

_add_special_tokens = False

get_size()

token_to_id(token)

get_pad_id()

__call__(text, truncation=True)

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

Parameters:

• 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).

• truncation (Optional[bool])

save(dir_path)

Parameters:

dir_path (str)

classmethod load(tokenizer_path, relcat_config, **kwargs)

Parameters:

• tokenizer_path (str)

• relcat_config (medcat.config.config_rel_cat.ConfigRelCAT)

Return type:

BaseTokenizerWrapper

_backends = ['tokenizers']

vocab_files_names

slow_tokenizer_class: transformers.tokenization_utils.PreTrainedTokenizer = None

add_prefix_space

_tokenizer

_decode_use_source_tokenizer = False

property is_fast: bool

Return type:

bool

property can_save_slow_tokenizer: bool

Whether or not the slow tokenizer can be saved. Usually for sentencepiece based slow tokenizer, this can only be True if the original “sentencepiece.model” was not deleted.

Type:

bool

Return type:

bool

property vocab_size: int

Size of the base vocabulary (without the added tokens).

Type:

int

Return type:

int

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:

`Dict[str, int]` – The vocabulary.

Return type:

Dict[str, int]

property vocab: Dict[str, int]

Return type:

Dict[str, int]

property added_tokens_encoder: Dict[str, int]

Returns the sorted mapping from string to index. The added tokens encoder is cached for performance optimisation in self._added_tokens_encoder for the slow tokenizers.

Return type:

Dict[str, int]

property added_tokens_decoder: Dict[int, transformers.tokenization_utils_base.AddedToken]

Returns the added tokens in the vocabulary as a dictionary of index to AddedToken.

Returns:

`Dict[str, int]` – The added tokens.

Return type:

Dict[int, transformers.tokenization_utils_base.AddedToken]

get_added_vocab()

Returns the added tokens in the vocabulary as a dictionary of token to index.

Returns:

`Dict[str, int]` – The added tokens.

Return type:

Dict[str, int]

__len__()

Size of the full vocabulary with the added tokens.

Return type:

int

property backend_tokenizer: tokenizers.Tokenizer

The Rust tokenizer used as a backend.

Type:

tokenizers.implementations.BaseTokenizer

Return type:

tokenizers.Tokenizer

property decoder: tokenizers.decoders.Decoder

The Rust decoder for this tokenizer.

Type:

tokenizers.decoders.Decoder

Return type:

tokenizers.decoders.Decoder

_convert_encoding(encoding, 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)

Convert the encoding representation (from low-level HuggingFace tokenizer output) to a python Dict and a list of encodings, take care of building a batch from overflowing tokens.

Overflowing tokens are converted to additional examples (like batches) so the output values of the dict are lists (overflows) of lists (tokens).

Output shape: (overflows, sequence length)

Parameters:

• encoding (tokenizers.Encoding)

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

Return type:

Tuple[Dict[str, Any], List[tokenizers.Encoding]]

convert_tokens_to_ids(tokens)

Converts a token string (or a sequence of tokens) in a single integer id (or a Iterable of ids), using the vocabulary.

Parameters:

tokens (str or Iterable[str]) – One or several token(s) to convert to token id(s).

Returns:

`int` or `List[int]` – The token id or list of token ids.

Return type:

Union[int, List[int]]

_convert_token_to_id_with_added_voc(token)

Parameters:

token (str)

Return type:

int

_convert_id_to_token(index)

Parameters:

index (int)

Return type:

Optional[str]

_add_tokens(new_tokens, special_tokens=False)

Parameters:

new_tokens (List[Union[str, transformers.tokenization_utils_base.AddedToken]])

Return type:

int

num_special_tokens_to_add(pair=False)

Returns the number of added tokens when encoding a sequence with special tokens.

<Tip>

This encodes a dummy input and checks the number of added tokens, and is therefore not efficient. Do not put this inside your training loop.

</Tip>

Parameters:

pair (bool, optional, defaults to False) – Whether the number of added tokens should be computed in the case of a sequence pair or a single sequence.

Returns:

`int` – Number of special tokens added to sequences.

Return type:

int

convert_ids_to_tokens(ids, skip_special_tokens=False)

Converts a single index or a sequence of indices in a token or a sequence of tokens, using the vocabulary and added tokens.

Parameters:

• ids (int or List[int]) – The token id (or token ids) to convert to tokens.

• skip_special_tokens (bool, optional, defaults to False) – Whether or not to remove special tokens in the decoding.

Returns:

`str` or `List[str]` – The decoded token(s).

Return type:

Union[str, List[str]]

tokenize(text, pair=None, add_special_tokens=False, **kwargs)

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

Parameters:

• 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.

Return type:

List[str]

set_truncation_and_padding(padding_strategy, truncation_strategy, max_length, stride, pad_to_multiple_of, padding_side)

Define the truncation and the padding strategies for fast tokenizers (provided by HuggingFace tokenizers library) and restore the tokenizer settings afterwards.

The provided tokenizer has no padding / truncation strategy before the managed section. If your tokenizer set a padding / truncation strategy before, then it will be reset to no padding / truncation when exiting the managed section.

Parameters:

• padding_strategy ([~utils.PaddingStrategy]) – The kind of padding that will be applied to the input

• truncation_strategy ([~tokenization_utils_base.TruncationStrategy]) – The kind of truncation that will be applied to the input

• max_length (int) – The maximum size of a sequence.

• stride (int) – The stride to use when handling overflow.

• 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).

• padding_side (str, optional) – The side on which the model should have padding applied. Should be selected between [‘right’, ‘left’]. Default value is picked from the class attribute of the same name.

_batch_encode_plus(batch_text_or_text_pairs, add_special_tokens=True, padding_strategy=PaddingStrategy.DO_NOT_PAD, truncation_strategy=TruncationStrategy.DO_NOT_TRUNCATE, max_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, padding_side=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)

Parameters:

• batch_text_or_text_pairs (Union[List[transformers.tokenization_utils_base.TextInput], List[transformers.tokenization_utils_base.TextInputPair], List[transformers.tokenization_utils_base.PreTokenizedInput], List[transformers.tokenization_utils_base.PreTokenizedInputPair]])

• add_special_tokens (bool)

• padding_strategy (transformers.utils.PaddingStrategy)

• truncation_strategy (transformers.tokenization_utils_base.TruncationStrategy)

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[str])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

transformers.tokenization_utils_base.BatchEncoding

_encode_plus(text, text_pair=None, add_special_tokens=True, padding_strategy=PaddingStrategy.DO_NOT_PAD, truncation_strategy=TruncationStrategy.DO_NOT_TRUNCATE, max_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, padding_side=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)

Parameters:

• text (Union[transformers.tokenization_utils_base.TextInput, transformers.tokenization_utils_base.PreTokenizedInput])

• text_pair (Optional[Union[transformers.tokenization_utils_base.TextInput, transformers.tokenization_utils_base.PreTokenizedInput]])

• add_special_tokens (bool)

• padding_strategy (transformers.utils.PaddingStrategy)

• truncation_strategy (transformers.tokenization_utils_base.TruncationStrategy)

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[bool])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

transformers.tokenization_utils_base.BatchEncoding

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.

Parameters:

tokens (List[str]) – The token to join in a string.

Returns:

`str` – The joined tokens.

Return type:

str

_decode(token_ids, skip_special_tokens=False, clean_up_tokenization_spaces=None, **kwargs)

Parameters:

• token_ids (Union[int, List[int]])

• skip_special_tokens (bool)

• clean_up_tokenization_spaces (bool)

Return type:

str

_save_pretrained(save_directory, file_names, legacy_format=None, filename_prefix=None)

Save a tokenizer using the slow-tokenizer/legacy format: vocabulary + added tokens as well as in a unique JSON file containing {config + vocab + added-tokens}.

Parameters:

• save_directory (Union[str, os.PathLike])

• file_names (Tuple[str])

• legacy_format (Optional[bool])

• filename_prefix (Optional[str])

Return type:

Tuple[str]

train_new_from_iterator(text_iterator, vocab_size, length=None, new_special_tokens=None, special_tokens_map=None, **kwargs)

Trains a tokenizer on a new corpus with the same defaults (in terms of special tokens or tokenization pipeline) as the current one.

Parameters:

• text_iterator (generator of List[str]) – The training corpus. Should be a generator of batches of texts, for instance a list of lists of texts if you have everything in memory.

• vocab_size (int) – The size of the vocabulary you want for your tokenizer.

• length (int, optional) – The total number of sequences in the iterator. This is used to provide meaningful progress tracking

• new_special_tokens (list of str or AddedToken, optional) – A list of new special tokens to add to the tokenizer you are training.

• special_tokens_map (Dict[str, str], optional) – If you want to rename some of the special tokens this tokenizer uses, pass along a mapping old special token name to new special token name in this argument.

• kwargs (Dict[str, Any], optional) – Additional keyword arguments passed along to the trainer from the 🤗 Tokenizers library.

Returns:

• [`PreTrainedTokenizerFast`] – A new tokenizer of the same type as the original one, trained on

• `text_iterator`.

pretrained_vocab_files_map: Dict[str, Dict[str, str]]

_auto_class: str | None = None

model_input_names: List[str] = ['input_ids', 'token_type_ids', 'attention_mask']

padding_side: str = 'right'

truncation_side: str = 'right'

init_inputs = ()

init_kwargs

name_or_path

_processor_class

model_max_length

clean_up_tokenization_spaces

split_special_tokens

deprecation_warnings

_in_target_context_manager = False

chat_template

extra_special_tokens

property max_len_single_sentence: int

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

Type:

int

Return type:

int

property max_len_sentences_pair: int

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

Type:

int

Return type:

int

_set_processor_class(processor_class)

Sets processor class as an attribute.

Parameters:

processor_class (str)

__repr__()

Return type:

str

apply_chat_template(conversation, tools=None, documents=None, chat_template=None, add_generation_prompt=False, continue_final_message=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.

Parameters:

• 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) – If this is set, a prompt with the token(s) that indicate the start of an assistant message will be appended to the formatted output. 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.

• continue_final_message (bool, optional) – If this is set, the chat will be formatted so that the final message in the chat is open-ended, without any EOS tokens. The model will continue this message rather than starting a new one. This allows you to “prefill” part of the model’s response for it. Cannot be used at the same time as add_generation_prompt.

• 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 torch.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.

• (`Dict[str (tokenizer_kwargs) –

  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.

• tokenizer_kwargs (Optional[Dict[str, Any]])

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.

Return type:

Union[str, List[int], List[str], List[List[int]], BatchEncoding]

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.

Parameters:

• 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.

Return type:

str

classmethod 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)

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

Parameters:

• 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>” ```

classmethod _from_pretrained(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)

static _eventually_correct_t5_max_length(pretrained_model_name_or_path, max_model_length, init_max_model_length)

classmethod convert_added_tokens(obj, save=False, add_type_field=True)

Parameters:

obj (Union[tokenizers.AddedToken, Any])

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).

Parameters:

• 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.

Return type:

Tuple[str]

abstract 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.

Parameters:

• 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.

Return type:

Tuple[str]

encode(text, text_pair=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, stride=0, padding_side=None, 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)).

Parameters:

• 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).

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

Return type:

List[int]

_get_padding_truncation_strategies(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.

_call_one(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, padding_side=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)

Parameters:

• text (Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]])

• text_pair (Optional[Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]]])

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

BatchEncoding

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, padding_side=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.

<Tip warning={true}>

This method is deprecated, __call__ should be used instead.

</Tip>

Parameters:

• 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).

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

Return type:

BatchEncoding

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, padding_side=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.

<Tip warning={true}>

This method is deprecated, __call__ should be used instead.

</Tip>

Parameters:

• 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).

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

BatchEncoding

pad(encoded_inputs, padding=True, max_length=None, pad_to_multiple_of=None, padding_side=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.

<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>

Parameters:

• 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).

• padding_side (str, optional) – The side on which the model should have padding applied. Should be selected between [‘right’, ‘left’]. Default value is picked from the class attribute of the same name.

• 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 torch.Tensor objects.

  • ’np’: Return Numpy np.ndarray objects.

• verbose (bool, optional, defaults to True) – Whether or not to print more information and warnings.

Return type:

BatchEncoding

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?](../glossary#token-type-ids)

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

Parameters:

• 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.

Return type:

List[int]

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.

Parameters:

• 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.

Return type:

List[int]

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, padding_side=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.

Parameters:

• 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.

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• prepend_batch_axis (bool)

Return type:

BatchEncoding

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.

Parameters:

• 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)

Return type:

Tuple[List[int], List[int], List[int]]

_pad(encoded_inputs, max_length=None, padding_strategy=PaddingStrategy.DO_NOT_PAD, pad_to_multiple_of=None, padding_side=None, return_attention_mask=None)

Pad encoded inputs (on left/right and up to predefined length or max length in the batch)

Parameters:

• encoded_inputs (Union[Dict[str, EncodedInput], BatchEncoding]) – Dictionary of tokenized inputs (List[int]) or batch of tokenized inputs (List[List[int]]).

• max_length (Optional[int]) – maximum length of the returned list and optionally padding length (see below). Will truncate by taking into account the special tokens.

• padding_strategy (transformers.utils.PaddingStrategy) –

  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 padding_side argument:

  • ’left’: pads on the left of the sequences

  • ’right’: pads on the right of the sequences

• pad_to_multiple_of (Optional[int]) – (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).

• padding_side (Optional[bool]) – The side on which the model should have padding applied. Should be selected between [‘right’, ‘left’]. Default value is picked from the class attribute of the same name.

• return_attention_mask (Optional[bool]) – (optional) Set to False to avoid returning attention mask (default: set to model specifics)

Return type:

dict

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.

Parameters:

• sequences (Union[List[int], List[List[int]], np.ndarray, torch.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 type:

List[str]

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)).

Parameters:

• token_ids (Union[int, List[int], np.ndarray, torch.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.

Return type:

str

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.

Parameters:

• 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.

Return type:

List[int]

static clean_up_tokenization(out_string)

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

Parameters:

out_string (str) – The text to clean up.

Returns:

`str` – The cleaned-up string.

Return type:

str

_eventual_warn_about_too_long_sequence(ids, max_length, verbose)

Depending on the input and internal state we might trigger a warning about a sequence that is too long for its corresponding model

Parameters:

• 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.

_switch_to_input_mode()

Private method to put the tokenizer in input mode (when it has different modes for input/outputs)

_switch_to_target_mode()

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

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.

classmethod register_for_auto_class(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>

Parameters:

auto_class (str or type, optional, defaults to “AutoTokenizer”) – The auto class to register this new tokenizer with.

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.

Parameters:

• 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 torch.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__.

Returns:

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.

Return type:

[BatchEncoding]

SPECIAL_TOKENS_ATTRIBUTES = ['bos_token', 'eos_token', 'unk_token', 'sep_token', 'pad_token', 'cls_token', 'mask_token',...

_pad_token_type_id = 0

verbose = False

_special_tokens_map

sanitize_special_tokens()

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

Return type:

int

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>’).

Parameters:

• 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._special_tokens_map[“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.

Return type:

int

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>” ```

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 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.

Parameters:

• 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.

Return type:

int

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)) ```

property pad_token_type_id: int

Id of the padding token type in the vocabulary.

Type:

int

Return type:

int

__setattr__(key, value)

__getattr__(key)

property special_tokens_map: Dict[str, 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.

Type:

Dict[str, Union[str, List[str]]]

Return type:

Dict[str, Union[str, List[str]]]

property special_tokens_map_extended: Dict[str, str | tokenizers.AddedToken | List[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.

Type:

Dict[str, Union[str, tokenizers.AddedToken, List[Union[str, tokenizers.AddedToken]]]]

Return type:

Dict[str, Union[str, tokenizers.AddedToken, List[Union[str, tokenizers.AddedToken]]]]

property all_special_tokens_extended: List[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.

Type:

List[Union[str, tokenizers.AddedToken]]

Return type:

List[Union[str, tokenizers.AddedToken]]

property all_special_tokens: List[str]

A list of the unique special tokens (‘<unk>’, ‘<cls>’, …, etc.).

Convert tokens of tokenizers.AddedToken type to string.

Type:

List[str]

Return type:

List[str]

property all_special_ids: List[int]

List the ids of the special tokens(‘<unk>’, ‘<cls>’, etc.) mapped to class attributes.

Type:

List[int]

Return type:

List[int]

_set_model_specific_special_tokens(special_tokens)

Adds new special tokens to the “SPECIAL_TOKENS_ATTRIBUTES” list which will be part of “self.special_tokens” and saved as a special token in tokenizer’s config. This allows us to dynamically add new model-type specific tokens after initilizing the tokenizer. For example: if the model tokenizers is multimodal, we can support special image or audio tokens.

Parameters:

special_tokens (List[str])

_create_repo(repo_id, private=None, token=None, repo_url=None, organization=None)

Create the repo if needed, cleans up repo_id with deprecated kwargs repo_url and organization, retrieves the token.

Parameters:

• repo_id (str)

• private (Optional[bool])

• token (Optional[Union[bool, str]])

• repo_url (Optional[str])

• organization (Optional[str])

Return type:

str

_get_files_timestamps(working_dir)

Returns the list of files with their last modification timestamp.

Parameters:

working_dir (Union[str, os.PathLike])

_upload_modified_files(working_dir, repo_id, files_timestamps, commit_message=None, token=None, create_pr=False, revision=None, commit_description=None)

Uploads all modified files in working_dir to repo_id, based on files_timestamps.

Parameters:

• working_dir (Union[str, os.PathLike])

• repo_id (str)

• files_timestamps (Dict[str, float])

• commit_message (Optional[str])

• token (Optional[Union[bool, str]])

• create_pr (bool)

• revision (str)

• commit_description (str)

push_to_hub(repo_id, use_temp_dir=None, commit_message=None, private=None, token=None, max_shard_size='5GB', create_pr=False, safe_serialization=True, revision=None, commit_description=None, tags=None, **deprecated_kwargs)

Upload the {object_files} to the 🤗 Model Hub.

Parameters:

• repo_id (str) – The name of the repository you want to push your {object} to. It should contain your organization name when pushing to a given organization.

• use_temp_dir (bool, optional) – Whether or not to use a temporary directory to store the files saved before they are pushed to the Hub. Will default to True if there is no directory named like repo_id, False otherwise.

• commit_message (str, optional) – Message to commit while pushing. Will default to “Upload {object}”.

• private (bool, optional) – Whether to make the repo private. If None (default), the repo will be public unless the organization’s default is private. This value is ignored if the repo already exists.

• token (bool or str, 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). Will default to True if repo_url is not specified.

• max_shard_size (int or str, optional, defaults to “5GB”) – Only applicable for models. The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size lower than this size. If expressed as a string, needs to be digits followed by a unit (like “5MB”). We default it to “5GB” so that users can easily load models on free-tier Google Colab instances without any CPU OOM issues.

• create_pr (bool, optional, defaults to False) – Whether or not to create a PR with the uploaded files or directly commit.

• safe_serialization (bool, optional, defaults to True) – Whether or not to convert the model weights in safetensors format for safer serialization.

• revision (str, optional) – Branch to push the uploaded files to.

• commit_description (str, optional) – The description of the commit that will be created

• tags (List[str], optional) – List of tags to push on the Hub.

Return type:

str

Examples:

```python from transformers import {object_class}

{object} = {object_class}.from_pretrained(“google-bert/bert-base-cased”)

# Push the {object} to your namespace with the name “my-finetuned-bert”. {object}.push_to_hub(“my-finetuned-bert”)

# Push the {object} to an organization with the name “my-finetuned-bert”. {object}.push_to_hub(“huggingface/my-finetuned-bert”) ```

medcat.components.addons.relation_extraction.llama.tokenizer.logger

class medcat.components.addons.relation_extraction.llama.tokenizer.TokenizerWrapperLlama(hf_tokenizers=None, max_seq_length=None, add_special_tokens=False)

Bases: medcat.components.addons.relation_extraction.tokenizer.BaseTokenizerWrapper

Wrapper around a huggingface Llama tokenizer so that it works with the RelCAT models.

Parameters:

• hf_tokenizers (transformers.LlamaTokenizerFast) – A huggingface Fast Llama.

• max_seq_length (Optional[int])

• add_special_tokens (Optional[bool])

name = 'tokenizer_wrapper_llama_rel'

pretrained_model_name_or_path = 'meta-llama/Llama-3.1-8B'

classmethod load(tokenizer_path, relcat_config, **kwargs)

Parameters:

• tokenizer_path (str)

• relcat_config (medcat.config.config_rel_cat.ConfigRelCAT)

Return type:

TokenizerWrapperLlama

__init__(hf_tokenizers=None, max_seq_length=None, add_special_tokens=False)

Parameters:

• max_seq_length (Optional[int])

• add_special_tokens (Optional[bool])

hf_tokenizers = None

max_seq_length = None

_add_special_tokens = False

get_size()

token_to_id(token)

get_pad_id()

__call__(text, truncation=True)

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

Parameters:

• 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).

• truncation (Optional[bool])

save(dir_path)

Parameters:

dir_path (str)

_backends = ['tokenizers']

vocab_files_names

slow_tokenizer_class: transformers.tokenization_utils.PreTrainedTokenizer = None

add_prefix_space

_tokenizer

_decode_use_source_tokenizer = False

property is_fast: bool

Return type:

bool

property can_save_slow_tokenizer: bool

Whether or not the slow tokenizer can be saved. Usually for sentencepiece based slow tokenizer, this can only be True if the original “sentencepiece.model” was not deleted.

Type:

bool

Return type:

bool

property vocab_size: int

Size of the base vocabulary (without the added tokens).

Type:

int

Return type:

int

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:

`Dict[str, int]` – The vocabulary.

Return type:

Dict[str, int]

property vocab: Dict[str, int]

Return type:

Dict[str, int]

property added_tokens_encoder: Dict[str, int]

Returns the sorted mapping from string to index. The added tokens encoder is cached for performance optimisation in self._added_tokens_encoder for the slow tokenizers.

Return type:

Dict[str, int]

property added_tokens_decoder: Dict[int, transformers.tokenization_utils_base.AddedToken]

Returns the added tokens in the vocabulary as a dictionary of index to AddedToken.

Returns:

`Dict[str, int]` – The added tokens.

Return type:

Dict[int, transformers.tokenization_utils_base.AddedToken]

get_added_vocab()

Returns the added tokens in the vocabulary as a dictionary of token to index.

Returns:

`Dict[str, int]` – The added tokens.

Return type:

Dict[str, int]

__len__()

Size of the full vocabulary with the added tokens.

Return type:

int

property backend_tokenizer: tokenizers.Tokenizer

The Rust tokenizer used as a backend.

Type:

tokenizers.implementations.BaseTokenizer

Return type:

tokenizers.Tokenizer

property decoder: tokenizers.decoders.Decoder

The Rust decoder for this tokenizer.

Type:

tokenizers.decoders.Decoder

Return type:

tokenizers.decoders.Decoder

_convert_encoding(encoding, 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)

Convert the encoding representation (from low-level HuggingFace tokenizer output) to a python Dict and a list of encodings, take care of building a batch from overflowing tokens.

Overflowing tokens are converted to additional examples (like batches) so the output values of the dict are lists (overflows) of lists (tokens).

Output shape: (overflows, sequence length)

Parameters:

• encoding (tokenizers.Encoding)

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

Return type:

Tuple[Dict[str, Any], List[tokenizers.Encoding]]

convert_tokens_to_ids(tokens)

Converts a token string (or a sequence of tokens) in a single integer id (or a Iterable of ids), using the vocabulary.

Parameters:

tokens (str or Iterable[str]) – One or several token(s) to convert to token id(s).

Returns:

`int` or `List[int]` – The token id or list of token ids.

Return type:

Union[int, List[int]]

_convert_token_to_id_with_added_voc(token)

Parameters:

token (str)

Return type:

int

_convert_id_to_token(index)

Parameters:

index (int)

Return type:

Optional[str]

_add_tokens(new_tokens, special_tokens=False)

Parameters:

new_tokens (List[Union[str, transformers.tokenization_utils_base.AddedToken]])

Return type:

int

num_special_tokens_to_add(pair=False)

Returns the number of added tokens when encoding a sequence with special tokens.

<Tip>

This encodes a dummy input and checks the number of added tokens, and is therefore not efficient. Do not put this inside your training loop.

</Tip>

Parameters:

pair (bool, optional, defaults to False) – Whether the number of added tokens should be computed in the case of a sequence pair or a single sequence.

Returns:

`int` – Number of special tokens added to sequences.

Return type:

int

convert_ids_to_tokens(ids, skip_special_tokens=False)

Converts a single index or a sequence of indices in a token or a sequence of tokens, using the vocabulary and added tokens.

Parameters:

• ids (int or List[int]) – The token id (or token ids) to convert to tokens.

• skip_special_tokens (bool, optional, defaults to False) – Whether or not to remove special tokens in the decoding.

Returns:

`str` or `List[str]` – The decoded token(s).

Return type:

Union[str, List[str]]

tokenize(text, pair=None, add_special_tokens=False, **kwargs)

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

Parameters:

• 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.

Return type:

List[str]

set_truncation_and_padding(padding_strategy, truncation_strategy, max_length, stride, pad_to_multiple_of, padding_side)

Define the truncation and the padding strategies for fast tokenizers (provided by HuggingFace tokenizers library) and restore the tokenizer settings afterwards.

The provided tokenizer has no padding / truncation strategy before the managed section. If your tokenizer set a padding / truncation strategy before, then it will be reset to no padding / truncation when exiting the managed section.

Parameters:

• padding_strategy ([~utils.PaddingStrategy]) – The kind of padding that will be applied to the input

• truncation_strategy ([~tokenization_utils_base.TruncationStrategy]) – The kind of truncation that will be applied to the input

• max_length (int) – The maximum size of a sequence.

• stride (int) – The stride to use when handling overflow.

• 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).

• padding_side (str, optional) – The side on which the model should have padding applied. Should be selected between [‘right’, ‘left’]. Default value is picked from the class attribute of the same name.

_batch_encode_plus(batch_text_or_text_pairs, add_special_tokens=True, padding_strategy=PaddingStrategy.DO_NOT_PAD, truncation_strategy=TruncationStrategy.DO_NOT_TRUNCATE, max_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, padding_side=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)

Parameters:

• batch_text_or_text_pairs (Union[List[transformers.tokenization_utils_base.TextInput], List[transformers.tokenization_utils_base.TextInputPair], List[transformers.tokenization_utils_base.PreTokenizedInput], List[transformers.tokenization_utils_base.PreTokenizedInputPair]])

• add_special_tokens (bool)

• padding_strategy (transformers.utils.PaddingStrategy)

• truncation_strategy (transformers.tokenization_utils_base.TruncationStrategy)

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[str])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

transformers.tokenization_utils_base.BatchEncoding

_encode_plus(text, text_pair=None, add_special_tokens=True, padding_strategy=PaddingStrategy.DO_NOT_PAD, truncation_strategy=TruncationStrategy.DO_NOT_TRUNCATE, max_length=None, stride=0, is_split_into_words=False, pad_to_multiple_of=None, padding_side=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)

Parameters:

• text (Union[transformers.tokenization_utils_base.TextInput, transformers.tokenization_utils_base.PreTokenizedInput])

• text_pair (Optional[Union[transformers.tokenization_utils_base.TextInput, transformers.tokenization_utils_base.PreTokenizedInput]])

• add_special_tokens (bool)

• padding_strategy (transformers.utils.PaddingStrategy)

• truncation_strategy (transformers.tokenization_utils_base.TruncationStrategy)

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[bool])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

transformers.tokenization_utils_base.BatchEncoding

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.

Parameters:

tokens (List[str]) – The token to join in a string.

Returns:

`str` – The joined tokens.

Return type:

str

_decode(token_ids, skip_special_tokens=False, clean_up_tokenization_spaces=None, **kwargs)

Parameters:

• token_ids (Union[int, List[int]])

• skip_special_tokens (bool)

• clean_up_tokenization_spaces (bool)

Return type:

str

_save_pretrained(save_directory, file_names, legacy_format=None, filename_prefix=None)

Save a tokenizer using the slow-tokenizer/legacy format: vocabulary + added tokens as well as in a unique JSON file containing {config + vocab + added-tokens}.

Parameters:

• save_directory (Union[str, os.PathLike])

• file_names (Tuple[str])

• legacy_format (Optional[bool])

• filename_prefix (Optional[str])

Return type:

Tuple[str]

train_new_from_iterator(text_iterator, vocab_size, length=None, new_special_tokens=None, special_tokens_map=None, **kwargs)

Trains a tokenizer on a new corpus with the same defaults (in terms of special tokens or tokenization pipeline) as the current one.

Parameters:

• text_iterator (generator of List[str]) – The training corpus. Should be a generator of batches of texts, for instance a list of lists of texts if you have everything in memory.

• vocab_size (int) – The size of the vocabulary you want for your tokenizer.

• length (int, optional) – The total number of sequences in the iterator. This is used to provide meaningful progress tracking

• new_special_tokens (list of str or AddedToken, optional) – A list of new special tokens to add to the tokenizer you are training.

• special_tokens_map (Dict[str, str], optional) – If you want to rename some of the special tokens this tokenizer uses, pass along a mapping old special token name to new special token name in this argument.

• kwargs (Dict[str, Any], optional) – Additional keyword arguments passed along to the trainer from the 🤗 Tokenizers library.

Returns:

• [`PreTrainedTokenizerFast`] – A new tokenizer of the same type as the original one, trained on

• `text_iterator`.

pretrained_vocab_files_map: Dict[str, Dict[str, str]]

_auto_class: str | None = None

model_input_names: List[str] = ['input_ids', 'token_type_ids', 'attention_mask']

padding_side: str = 'right'

truncation_side: str = 'right'

init_inputs = ()

init_kwargs

name_or_path

_processor_class

model_max_length

clean_up_tokenization_spaces

split_special_tokens

deprecation_warnings

_in_target_context_manager = False

chat_template

extra_special_tokens

property max_len_single_sentence: int

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

Type:

int

Return type:

int

property max_len_sentences_pair: int

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

Type:

int

Return type:

int

_set_processor_class(processor_class)

Sets processor class as an attribute.

Parameters:

processor_class (str)

__repr__()

Return type:

str

apply_chat_template(conversation, tools=None, documents=None, chat_template=None, add_generation_prompt=False, continue_final_message=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.

Parameters:

• 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) – If this is set, a prompt with the token(s) that indicate the start of an assistant message will be appended to the formatted output. 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.

• continue_final_message (bool, optional) – If this is set, the chat will be formatted so that the final message in the chat is open-ended, without any EOS tokens. The model will continue this message rather than starting a new one. This allows you to “prefill” part of the model’s response for it. Cannot be used at the same time as add_generation_prompt.

• 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 torch.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.

• (`Dict[str (tokenizer_kwargs) –

  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.

• tokenizer_kwargs (Optional[Dict[str, Any]])

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.

Return type:

Union[str, List[int], List[str], List[List[int]], BatchEncoding]

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.

Parameters:

• 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.

Return type:

str

classmethod 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)

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

Parameters:

• 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>” ```

classmethod _from_pretrained(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)

static _eventually_correct_t5_max_length(pretrained_model_name_or_path, max_model_length, init_max_model_length)

classmethod convert_added_tokens(obj, save=False, add_type_field=True)

Parameters:

obj (Union[tokenizers.AddedToken, Any])

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).

Parameters:

• 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.

Return type:

Tuple[str]

abstract 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.

Parameters:

• 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.

Return type:

Tuple[str]

encode(text, text_pair=None, add_special_tokens=True, padding=False, truncation=None, max_length=None, stride=0, padding_side=None, 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)).

Parameters:

• 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).

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

Return type:

List[int]

_get_padding_truncation_strategies(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.

_call_one(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, padding_side=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)

Parameters:

• text (Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]])

• text_pair (Optional[Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]]])

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

BatchEncoding

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, padding_side=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.

<Tip warning={true}>

This method is deprecated, __call__ should be used instead.

</Tip>

Parameters:

• 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).

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

Return type:

BatchEncoding

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, padding_side=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.

<Tip warning={true}>

This method is deprecated, __call__ should be used instead.

</Tip>

Parameters:

• 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).

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• is_split_into_words (bool)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• split_special_tokens (bool)

Return type:

BatchEncoding

pad(encoded_inputs, padding=True, max_length=None, pad_to_multiple_of=None, padding_side=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.

<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>

Parameters:

• 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).

• padding_side (str, optional) – The side on which the model should have padding applied. Should be selected between [‘right’, ‘left’]. Default value is picked from the class attribute of the same name.

• 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 torch.Tensor objects.

  • ’np’: Return Numpy np.ndarray objects.

• verbose (bool, optional, defaults to True) – Whether or not to print more information and warnings.

Return type:

BatchEncoding

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?](../glossary#token-type-ids)

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

Parameters:

• 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.

Return type:

List[int]

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.

Parameters:

• 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.

Return type:

List[int]

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, padding_side=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.

Parameters:

• 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.

• add_special_tokens (bool)

• padding (Union[bool, str, transformers.utils.PaddingStrategy])

• truncation (Union[bool, str, TruncationStrategy])

• max_length (Optional[int])

• stride (int)

• pad_to_multiple_of (Optional[int])

• padding_side (Optional[bool])

• return_tensors (Optional[Union[str, transformers.utils.TensorType]])

• return_token_type_ids (Optional[bool])

• return_attention_mask (Optional[bool])

• return_overflowing_tokens (bool)

• return_special_tokens_mask (bool)

• return_offsets_mapping (bool)

• return_length (bool)

• verbose (bool)

• prepend_batch_axis (bool)

Return type:

BatchEncoding

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.

Parameters:

• 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)

Return type:

Tuple[List[int], List[int], List[int]]

_pad(encoded_inputs, max_length=None, padding_strategy=PaddingStrategy.DO_NOT_PAD, pad_to_multiple_of=None, padding_side=None, return_attention_mask=None)

Pad encoded inputs (on left/right and up to predefined length or max length in the batch)

Parameters:

• encoded_inputs (Union[Dict[str, EncodedInput], BatchEncoding]) – Dictionary of tokenized inputs (List[int]) or batch of tokenized inputs (List[List[int]]).

• max_length (Optional[int]) – maximum length of the returned list and optionally padding length (see below). Will truncate by taking into account the special tokens.

• padding_strategy (transformers.utils.PaddingStrategy) –

  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 padding_side argument:

  • ’left’: pads on the left of the sequences

  • ’right’: pads on the right of the sequences

• pad_to_multiple_of (Optional[int]) – (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).

• padding_side (Optional[bool]) – The side on which the model should have padding applied. Should be selected between [‘right’, ‘left’]. Default value is picked from the class attribute of the same name.

• return_attention_mask (Optional[bool]) – (optional) Set to False to avoid returning attention mask (default: set to model specifics)

Return type:

dict

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.

Parameters:

• sequences (Union[List[int], List[List[int]], np.ndarray, torch.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 type:

List[str]

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)).

Parameters:

• token_ids (Union[int, List[int], np.ndarray, torch.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.

Return type:

str

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.

Parameters:

• 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.

Return type:

List[int]

static clean_up_tokenization(out_string)

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

Parameters:

out_string (str) – The text to clean up.

Returns:

`str` – The cleaned-up string.

Return type:

str

_eventual_warn_about_too_long_sequence(ids, max_length, verbose)

Depending on the input and internal state we might trigger a warning about a sequence that is too long for its corresponding model

Parameters:

• 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.

_switch_to_input_mode()

Private method to put the tokenizer in input mode (when it has different modes for input/outputs)

_switch_to_target_mode()

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

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.

classmethod register_for_auto_class(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>

Parameters:

auto_class (str or type, optional, defaults to “AutoTokenizer”) – The auto class to register this new tokenizer with.

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.

Parameters:

• 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 torch.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__.

Returns:

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.

Return type:

[BatchEncoding]

SPECIAL_TOKENS_ATTRIBUTES = ['bos_token', 'eos_token', 'unk_token', 'sep_token', 'pad_token', 'cls_token', 'mask_token',...

_pad_token_type_id = 0

verbose = False

_special_tokens_map

sanitize_special_tokens()

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

Return type:

int

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>’).

Parameters:

• 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._special_tokens_map[“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.

Return type:

int

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>” ```

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 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.

Parameters:

• 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.

Return type:

int

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)) ```

property pad_token_type_id: int

Id of the padding token type in the vocabulary.

Type:

int

Return type:

int

__setattr__(key, value)

__getattr__(key)

property special_tokens_map: Dict[str, 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.

Type:

Dict[str, Union[str, List[str]]]

Return type:

Dict[str, Union[str, List[str]]]

property special_tokens_map_extended: Dict[str, str | tokenizers.AddedToken | List[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.

Type:

Dict[str, Union[str, tokenizers.AddedToken, List[Union[str, tokenizers.AddedToken]]]]

Return type:

Dict[str, Union[str, tokenizers.AddedToken, List[Union[str, tokenizers.AddedToken]]]]

property all_special_tokens_extended: List[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.

Type:

List[Union[str, tokenizers.AddedToken]]

Return type:

List[Union[str, tokenizers.AddedToken]]

property all_special_tokens: List[str]

A list of the unique special tokens (‘<unk>’, ‘<cls>’, …, etc.).

Convert tokens of tokenizers.AddedToken type to string.

Type:

List[str]

Return type:

List[str]

property all_special_ids: List[int]

List the ids of the special tokens(‘<unk>’, ‘<cls>’, etc.) mapped to class attributes.

Type:

List[int]

Return type:

List[int]

_set_model_specific_special_tokens(special_tokens)

Adds new special tokens to the “SPECIAL_TOKENS_ATTRIBUTES” list which will be part of “self.special_tokens” and saved as a special token in tokenizer’s config. This allows us to dynamically add new model-type specific tokens after initilizing the tokenizer. For example: if the model tokenizers is multimodal, we can support special image or audio tokens.

Parameters:

special_tokens (List[str])

_create_repo(repo_id, private=None, token=None, repo_url=None, organization=None)

Create the repo if needed, cleans up repo_id with deprecated kwargs repo_url and organization, retrieves the token.

Parameters:

• repo_id (str)

• private (Optional[bool])

• token (Optional[Union[bool, str]])

• repo_url (Optional[str])

• organization (Optional[str])

Return type:

str

_get_files_timestamps(working_dir)

Returns the list of files with their last modification timestamp.

Parameters:

working_dir (Union[str, os.PathLike])

_upload_modified_files(working_dir, repo_id, files_timestamps, commit_message=None, token=None, create_pr=False, revision=None, commit_description=None)

Uploads all modified files in working_dir to repo_id, based on files_timestamps.

Parameters:

• working_dir (Union[str, os.PathLike])

• repo_id (str)

• files_timestamps (Dict[str, float])

• commit_message (Optional[str])

• token (Optional[Union[bool, str]])

• create_pr (bool)

• revision (str)

• commit_description (str)

push_to_hub(repo_id, use_temp_dir=None, commit_message=None, private=None, token=None, max_shard_size='5GB', create_pr=False, safe_serialization=True, revision=None, commit_description=None, tags=None, **deprecated_kwargs)

Upload the {object_files} to the 🤗 Model Hub.

Parameters:

• repo_id (str) – The name of the repository you want to push your {object} to. It should contain your organization name when pushing to a given organization.

• use_temp_dir (bool, optional) – Whether or not to use a temporary directory to store the files saved before they are pushed to the Hub. Will default to True if there is no directory named like repo_id, False otherwise.

• commit_message (str, optional) – Message to commit while pushing. Will default to “Upload {object}”.

• private (bool, optional) – Whether to make the repo private. If None (default), the repo will be public unless the organization’s default is private. This value is ignored if the repo already exists.

• token (bool or str, 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). Will default to True if repo_url is not specified.

• max_shard_size (int or str, optional, defaults to “5GB”) – Only applicable for models. The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size lower than this size. If expressed as a string, needs to be digits followed by a unit (like “5MB”). We default it to “5GB” so that users can easily load models on free-tier Google Colab instances without any CPU OOM issues.

• create_pr (bool, optional, defaults to False) – Whether or not to create a PR with the uploaded files or directly commit.

• safe_serialization (bool, optional, defaults to True) – Whether or not to convert the model weights in safetensors format for safer serialization.

• revision (str, optional) – Branch to push the uploaded files to.

• commit_description (str, optional) – The description of the commit that will be created

• tags (List[str], optional) – List of tags to push on the Hub.

Return type:

str

Examples:

```python from transformers import {object_class}

{object} = {object_class}.from_pretrained(“google-bert/bert-base-cased”)

# Push the {object} to your namespace with the name “my-finetuned-bert”. {object}.push_to_hub(“my-finetuned-bert”)

# Push the {object} to an organization with the name “my-finetuned-bert”. {object}.push_to_hub(“huggingface/my-finetuned-bert”) ```