data_collator.py

# Copyright 2020 The HuggingFace Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import random
import warnings
from dataclasses import dataclass
from typing import Any, Callable, Dict, List, NewType, Optional, Tuple, Union

import torch
from torch.nn.utils.rnn import pad_sequence

from ..file_utils import PaddingStrategy
from ..modeling_utils import PreTrainedModel
from ..tokenization_utils_base import BatchEncoding, PreTrainedTokenizerBase


InputDataClass = NewType("InputDataClass", Any)

"""
A DataCollator is a function that takes a list of samples from a Dataset and collate them into a batch, as a dictionary
of Tensors.
"""
DataCollator = NewType("DataCollator", Callable[[List[InputDataClass]], Dict[str, torch.Tensor]])


def default_data_collator(features: List[InputDataClass]) -> Dict[str, torch.Tensor]:
    """
    Very simple data collator that simply collates batches of dict-like objects and performs special handling for
    potential keys named:

        - ``label``: handles a single value (int or float) per object
        - ``label_ids``: handles a list of values per object

    Does not do any additional preprocessing: property names of the input object will be used as corresponding inputs
    to the model. See glue and ner for example of how it's useful.
    """

    # In this function we'll make the assumption that all `features` in the batch
    # have the same attributes.
    # So we will look at the first element as a proxy for what attributes exist
    # on the whole batch.
    if not isinstance(features[0], (dict, BatchEncoding)):
        features = [vars(f) for f in features]

    first = features[0]
    batch = {}

    # Special handling for labels.
    # Ensure that tensor is created with the correct type
    # (it should be automatically the case, but let's make sure of it.)
    if "label" in first and first["label"] is not None:
        label = first["label"].item() if isinstance(first["label"], torch.Tensor) else first["label"]
        dtype = torch.long if isinstance(label, int) else torch.float
        batch["labels"] = torch.tensor([f["label"] for f in features], dtype=dtype)
    elif "label_ids" in first and first["label_ids"] is not None:
        if isinstance(first["label_ids"], torch.Tensor):
            batch["labels"] = torch.stack([f["label_ids"] for f in features])
        else:
            dtype = torch.long if type(first["label_ids"][0]) is int else torch.float
            batch["labels"] = torch.tensor([f["label_ids"] for f in features], dtype=dtype)

    # Handling of all other possible keys.
    # Again, we will use the first element to figure out which key/values are not None for this model.
    for k, v in first.items():
        if k not in ("label", "label_ids") and v is not None and not isinstance(v, str):
            if isinstance(v, torch.Tensor):
                batch[k] = torch.stack([f[k] for f in features])
            else:
                batch[k] = torch.tensor([f[k] for f in features])

    return batch


@dataclass
class DataCollatorWithPadding:
    """
    Data collator that will dynamically pad the inputs received.

    Args:
        tokenizer (:class:`~transformers.PreTrainedTokenizer` or :class:`~transformers.PreTrainedTokenizerFast`):
            The tokenizer used for encoding the data.
        padding (:obj:`bool`, :obj:`str` or :class:`~transformers.file_utils.PaddingStrategy`, `optional`, defaults to :obj:`True`):
            Select a strategy to pad the returned sequences (according to the model's padding side and padding index)
            among:

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

    tokenizer: PreTrainedTokenizerBase
    padding: Union[bool, str, PaddingStrategy] = True
    max_length: Optional[int] = None
    pad_to_multiple_of: Optional[int] = None

    def __call__(self, features: List[Dict[str, Union[List[int], torch.Tensor]]]) -> Dict[str, torch.Tensor]:
        batch = self.tokenizer.pad(
            features,
            padding=self.padding,
            max_length=self.max_length,
            pad_to_multiple_of=self.pad_to_multiple_of,
            return_tensors="pt",
        )
        if "label" in batch:
            batch["labels"] = batch["label"]
            del batch["label"]
        if "label_ids" in batch:
            batch["labels"] = batch["label_ids"]
            del batch["label_ids"]
        return batch


@dataclass
class DataCollatorForTokenClassification:
    """
    Data collator that will dynamically pad the inputs received, as well as the labels.

    Args:
        tokenizer (:class:`~transformers.PreTrainedTokenizer` or :class:`~transformers.PreTrainedTokenizerFast`):
            The tokenizer used for encoding the data.
        padding (:obj:`bool`, :obj:`str` or :class:`~transformers.file_utils.PaddingStrategy`, `optional`, defaults to :obj:`True`):
            Select a strategy to pad the returned sequences (according to the model's padding side and padding index)
            among:

            * :obj:`True` or :obj:`'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
              sequence if provided).
            * :obj:`'max_length'`: Pad to a maximum length specified with the argument :obj:`max_length` or to the
              maximum acceptable input length for the model if that argument is not provided.
            * :obj:`False` or :obj:`'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of
              different lengths).
        max_length (:obj:`int`, `optional`):
            Maximum length of the returned list and optionally padding length (see above).
        pad_to_multiple_of (:obj:`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).
        label_pad_token_id (:obj:`int`, `optional`, defaults to -100):
            The id to use when padding the labels (-100 will be automatically ignore by PyTorch loss functions).
    """

    tokenizer: PreTrainedTokenizerBase
    padding: Union[bool, str, PaddingStrategy] = True
    max_length: Optional[int] = None
    pad_to_multiple_of: Optional[int] = None
    label_pad_token_id: int = -100

    def __call__(self, features):
        label_name = "label" if "label" in features[0].keys() else "labels"
        labels = [feature[label_name] for feature in features] if label_name in features[0].keys() else None
        batch = self.tokenizer.pad(
            features,
            padding=self.padding,
            max_length=self.max_length,
            pad_to_multiple_of=self.pad_to_multiple_of,
            # Conversion to tensors will fail if we have labels as they are not of the same length yet.
            return_tensors="pt" if labels is None else None,
        )

        if labels is None:
            return batch

        sequence_length = torch.tensor(batch["input_ids"]).shape[1]
        padding_side = self.tokenizer.padding_side
        if padding_side == "right":
            batch["labels"] = [label + [self.label_pad_token_id] * (sequence_length - len(label)) for label in labels]
        else:
            batch["labels"] = [[self.label_pad_token_id] * (sequence_length - len(label)) + label for label in labels]

        batch = {k: torch.tensor(v, dtype=torch.int64) for k, v in batch.items()}
        return batch


def _collate_batch(examples, tokenizer, pad_to_multiple_of: Optional[int] = None):
    """Collate `examples` into a batch, using the information in `tokenizer` for padding if necessary."""
    # Tensorize if necessary.
    if isinstance(examples[0], (list, tuple)):
        examples = [torch.tensor(e, dtype=torch.long) for e in examples]

    # Check if padding is necessary.
    length_of_first = examples[0].size(0)
    are_tensors_same_length = all(x.size(0) == length_of_first for x in examples)
    if are_tensors_same_length and (pad_to_multiple_of is None or length_of_first % pad_to_multiple_of == 0):
        return torch.stack(examples, dim=0)

    # If yes, check if we have a `pad_token`.
    if tokenizer._pad_token is None:
        raise ValueError(
            "You are attempting to pad samples but the tokenizer you are using"
            f" ({tokenizer.__class__.__name__}) does not have a pad token."
        )

    # Creating the full tensor and filling it with our data.
    max_length = max(x.size(0) for x in examples)
    if pad_to_multiple_of is not None and (max_length % pad_to_multiple_of != 0):
        max_length = ((max_length // pad_to_multiple_of) + 1) * pad_to_multiple_of
    result = examples[0].new_full([len(examples), max_length], tokenizer.pad_token_id)
    for i, example in enumerate(examples):
        if tokenizer.padding_side == "right":
            result[i, : example.shape[0]] = example
        else:
            result[i, -example.shape[0] :] = example
    return result


def tolist(x: Union[List[Any], torch.Tensor]):
    return x.tolist() if isinstance(x, torch.Tensor) else x


@dataclass
class DataCollatorForSeq2Seq:
    """
    Data collator that will dynamically pad the inputs received, as well as the labels.

    Args:
        tokenizer (:class:`~transformers.PreTrainedTokenizer` or :class:`~transformers.PreTrainedTokenizerFast`):
            The tokenizer used for encoding the data.
        model (:class:`~transformers.PreTrainedModel`):
            The model that is being trained. If set and has the `prepare_decoder_input_ids_from_labels`, use it to
            prepare the `decoder_input_ids`

            This is useful when using `label_smoothing` to avoid calculating loss twice.
        padding (:obj:`bool`, :obj:`str` or :class:`~transformers.file_utils.PaddingStrategy`, `optional`, defaults to :obj:`True`):
            Select a strategy to pad the returned sequences (according to the model's padding side and padding index)
            among:

            * :obj:`True` or :obj:`'longest'`: Pad to the longest sequence in the batch (or no padding if only a single
              sequence is provided).
            * :obj:`'max_length'`: Pad to a maximum length specified with the argument :obj:`max_length` or to the
              maximum acceptable input length for the model if that argument is not provided.
            * :obj:`False` or :obj:`'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of
              different lengths).
        max_length (:obj:`int`, `optional`):
            Maximum length of the returned list and optionally padding length (see above).
        pad_to_multiple_of (:obj:`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).
        label_pad_token_id (:obj:`int`, `optional`, defaults to -100):
            The id to use when padding the labels (-100 will be automatically ignored by PyTorch loss functions).
    """

    tokenizer: PreTrainedTokenizerBase
    model: Optional[PreTrainedModel] = None
    padding: Union[bool, str, PaddingStrategy] = True
    max_length: Optional[int] = None
    pad_to_multiple_of: Optional[int] = None
    label_pad_token_id: int = -100

    def __call__(self, features):
        labels = [feature["labels"] for feature in features] if "labels" in features[0].keys() else None
        # We have to pad the labels before calling `tokenizer.pad` as this method won't pad them and needs them of the
        # same length to return tensors.
        if labels is not None:
            max_label_length = max(len(l) for l in labels)
            padding_side = self.tokenizer.padding_side
            for feature in features:
                remainder = [self.label_pad_token_id] * (max_label_length - len(feature["labels"]))
                feature["labels"] = (
                    feature["labels"] + remainder if padding_side == "right" else remainder + feature["labels"]
                )

        features = self.tokenizer.pad(
            features,
            padding=self.padding,
            max_length=self.max_length,
            pad_to_multiple_of=self.pad_to_multiple_of,
            return_tensors="pt",
        )

        # prepare decoder_input_ids
        if self.model is not None and hasattr(self.model, "prepare_decoder_input_ids_from_labels"):
            decoder_input_ids = self.model.prepare_decoder_input_ids_from_labels(labels=features["labels"])
            features["decoder_input_ids"] = decoder_input_ids

        return features


@dataclass
class DataCollatorForLanguageModeling:
    """
    Data collator used for language modeling. Inputs are dynamically padded to the maximum length of a batch if they
    are not all of the same length.

    Args:
        tokenizer (:class:`~transformers.PreTrainedTokenizer` or :class:`~transformers.PreTrainedTokenizerFast`):
            The tokenizer used for encoding the data.
        mlm (:obj:`bool`, `optional`, defaults to :obj:`True`):
            Whether or not to use masked language modeling. If set to :obj:`False`, the labels are the same as the
            inputs with the padding tokens ignored (by setting them to -100). Otherwise, the labels are -100 for
            non-masked tokens and the value to predict for the masked token.
        mlm_probability (:obj:`float`, `optional`, defaults to 0.15):
            The probability with which to (randomly) mask tokens in the input, when :obj:`mlm` is set to :obj:`True`.
        pad_to_multiple_of (:obj:`int`, `optional`):
            If set will pad the sequence to a multiple of the provided value.

    .. note::

        For best performance, this data collator should be used with a dataset having items that are dictionaries or
        BatchEncoding, with the :obj:`"special_tokens_mask"` key, as returned by a
        :class:`~transformers.PreTrainedTokenizer` or a :class:`~transformers.PreTrainedTokenizerFast` with the
        argument :obj:`return_special_tokens_mask=True`.
    """

    tokenizer: PreTrainedTokenizerBase
    mlm: bool = True
    mlm_probability: float = 0.15
    pad_to_multiple_of: Optional[int] = None

    def __post_init__(self):
        if self.mlm and self.tokenizer.mask_token is None:
            raise ValueError(
                "This tokenizer does not have a mask token which is necessary for masked language modeling. "
                "You should pass `mlm=False` to train on causal language modeling instead."
            )

    def __call__(
        self, examples: List[Union[List[int], torch.Tensor, Dict[str, torch.Tensor]]]
    ) -> Dict[str, torch.Tensor]:
        # Handle dict or lists with proper padding and conversion to tensor.
        if isinstance(examples[0], (dict, BatchEncoding)):
            batch = self.tokenizer.pad(examples, return_tensors="pt", pad_to_multiple_of=self.pad_to_multiple_of)
        else:
            batch = {"input_ids": _collate_batch(examples, self.tokenizer, pad_to_multiple_of=self.pad_to_multiple_of)}

        # If special token mask has been preprocessed, pop it from the dict.
        special_tokens_mask = batch.pop("special_tokens_mask", None)
        if self.mlm:
            batch["input_ids"], batch["labels"] = self.mask_tokens(
                batch["input_ids"], special_tokens_mask=special_tokens_mask
            )
        else:
            labels = batch["input_ids"].clone()
            if self.tokenizer.pad_token_id is not None:
                labels[labels == self.tokenizer.pad_token_id] = -100
            batch["labels"] = labels
        return batch

    def mask_tokens(
        self, inputs: torch.Tensor, special_tokens_mask: Optional[torch.Tensor] = None
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        """
        Prepare masked tokens inputs/labels for masked language modeling: 80% MASK, 10% random, 10% original.
        """
        labels = inputs.clone()
        # We sample a few tokens in each sequence for MLM training (with probability `self.mlm_probability`)
        probability_matrix = torch.full(labels.shape, self.mlm_probability)
        if special_tokens_mask is None:
            special_tokens_mask = [
                self.tokenizer.get_special_tokens_mask(val, already_has_special_tokens=True) for val in labels.tolist()
            ]
            special_tokens_mask = torch.tensor(special_tokens_mask, dtype=torch.bool)
        else:
            special_tokens_mask = special_tokens_mask.bool()

        probability_matrix.masked_fill_(special_tokens_mask, value=0.0)
        masked_indices = torch.bernoulli(probability_matrix).bool()
        labels[~masked_indices] = -100  # We only compute loss on masked tokens

        # 80% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
        indices_replaced = torch.bernoulli(torch.full(labels.shape, 0.8)).bool() & masked_indices
        inputs[indices_replaced] = self.tokenizer.convert_tokens_to_ids(self.tokenizer.mask_token)

        # 10% of the time, we replace masked input tokens with random word
        indices_random = torch.bernoulli(torch.full(labels.shape, 0.5)).bool() & masked_indices & ~indices_replaced
        random_words = torch.randint(len(self.tokenizer), labels.shape, dtype=torch.long)
        inputs[indices_random] = random_words[indices_random]

        # The rest of the time (10% of the time) we keep the masked input tokens unchanged
        return inputs, labels


@dataclass
class DataCollatorForWholeWordMask(DataCollatorForLanguageModeling):
    """
    Data collator used for language modeling.

    - collates batches of tensors, honoring their tokenizer's pad_token
    - preprocesses batches for masked language modeling
    """

    def __call__(
        self, examples: List[Union[List[int], torch.Tensor, Dict[str, torch.Tensor]]]
    ) -> Dict[str, torch.Tensor]:
        if isinstance(examples[0], (dict, BatchEncoding)):
            input_ids = [e["input_ids"] for e in examples]
        else:
            input_ids = examples
            examples = [{"input_ids": e} for e in examples]

        batch_input = _collate_batch(input_ids, self.tokenizer)

        mask_labels = []
        for e in examples:
            ref_tokens = []
            for id in tolist(e["input_ids"]):
                token = self.tokenizer._convert_id_to_token(id)
                ref_tokens.append(token)

            # For Chinese tokens, we need extra inf to mark sub-word, e.g [喜,欢]-> [喜，##欢]
            if "chinese_ref" in e:
                ref_pos = tolist(e["chinese_ref"])
                len_seq = len(e["input_ids"])
                for i in range(len_seq):
                    if i in ref_pos:
                        ref_tokens[i] = "##" + ref_tokens[i]
            mask_labels.append(self._whole_word_mask(ref_tokens))
        batch_mask = _collate_batch(mask_labels, self.tokenizer)
        inputs, labels = self.mask_tokens(batch_input, batch_mask)
        return {"input_ids": inputs, "labels": labels}

    def _whole_word_mask(self, input_tokens: List[str], max_predictions=512):
        """
        Get 0/1 labels for masked tokens with whole word mask proxy
        """

        cand_indexes = []
        for (i, token) in enumerate(input_tokens):
            if token == "[CLS]" or token == "[SEP]":
                continue

            if len(cand_indexes) >= 1 and token.startswith("##"):
                cand_indexes[-1].append(i)
            else:
                cand_indexes.append([i])

        random.shuffle(cand_indexes)
        num_to_predict = min(max_predictions, max(1, int(round(len(input_tokens) * self.mlm_probability))))
        masked_lms = []
        covered_indexes = set()
        for index_set in cand_indexes:
            if len(masked_lms) >= num_to_predict:
                break
            # If adding a whole-word mask would exceed the maximum number of
            # predictions, then just skip this candidate.
            if len(masked_lms) + len(index_set) > num_to_predict:
                continue
            is_any_index_covered = False
            for index in index_set:
                if index in covered_indexes:
                    is_any_index_covered = True
                    break
            if is_any_index_covered:
                continue
            for index in index_set:
                covered_indexes.add(index)
                masked_lms.append(index)

        assert len(covered_indexes) == len(masked_lms)
        mask_labels = [1 if i in covered_indexes else 0 for i in range(len(input_tokens))]
        return mask_labels

    def mask_tokens(self, inputs: torch.Tensor, mask_labels: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor]:
        """
        Prepare masked tokens inputs/labels for masked language modeling: 80% MASK, 10% random, 10% original. Set
        'mask_labels' means we use whole word mask (wwm), we directly mask idxs according to it's ref.
        """

        if self.tokenizer.mask_token is None:
            raise ValueError(
                "This tokenizer does not have a mask token which is necessary for masked language modeling. Remove the --mlm flag if you want to use this tokenizer."
            )
        labels = inputs.clone()
        # We sample a few tokens in each sequence for masked-LM training (with probability args.mlm_probability defaults to 0.15 in Bert/RoBERTa)

        probability_matrix = mask_labels

        special_tokens_mask = [
            self.tokenizer.get_special_tokens_mask(val, already_has_special_tokens=True) for val in labels.tolist()
        ]
        probability_matrix.masked_fill_(torch.tensor(special_tokens_mask, dtype=torch.bool), value=0.0)
        if self.tokenizer._pad_token is not None:
            padding_mask = labels.eq(self.tokenizer.pad_token_id)
            probability_matrix.masked_fill_(padding_mask, value=0.0)

        masked_indices = probability_matrix.bool()
        labels[~masked_indices] = -100  # We only compute loss on masked tokens

        # 80% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
        indices_replaced = torch.bernoulli(torch.full(labels.shape, 0.8)).bool() & masked_indices
        inputs[indices_replaced] = self.tokenizer.convert_tokens_to_ids(self.tokenizer.mask_token)

        # 10% of the time, we replace masked input tokens with random word
        indices_random = torch.bernoulli(torch.full(labels.shape, 0.5)).bool() & masked_indices & ~indices_replaced
        random_words = torch.randint(len(self.tokenizer), labels.shape, dtype=torch.long)
        inputs[indices_random] = random_words[indices_random]

        # The rest of the time (10% of the time) we keep the masked input tokens unchanged
        return inputs, labels


@dataclass
class DataCollatorForSOP(DataCollatorForLanguageModeling):
    """
    Data collator used for sentence order prediction task.

    - collates batches of tensors, honoring their tokenizer's pad_token
    - preprocesses batches for both masked language modeling and sentence order prediction
    """

    def __init__(self, *args, **kwargs):
        warnings.warn(
            "DataCollatorForSOP is deprecated and will be removed in a future version, you can now use "
            "DataCollatorForLanguageModeling instead.",
            FutureWarning,
        )

    def __call__(self, examples: List[Dict[str, torch.Tensor]]) -> Dict[str, torch.Tensor]:
        input_ids = [example["input_ids"] for example in examples]
        input_ids = _collate_batch(input_ids, self.tokenizer)
        input_ids, labels, attention_mask = self.mask_tokens(input_ids)

        token_type_ids = [example["token_type_ids"] for example in examples]
        # size of segment_ids varied because randomness, padding zero to the end as the original implementation
        token_type_ids = pad_sequence(token_type_ids, batch_first=True, padding_value=self.tokenizer.pad_token_id)

        sop_label_list = [example["sentence_order_label"] for example in examples]
        sentence_order_label = torch.stack(sop_label_list)

        return {
            "input_ids": input_ids,
            "labels": labels,
            "attention_mask": attention_mask,
            "token_type_ids": token_type_ids,
            "sentence_order_label": sentence_order_label,
        }

    def mask_tokens(self, inputs: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
        """
        Prepare masked tokens inputs/labels/attention_mask for masked language modeling: 80% MASK, 10% random, 10%
        original. N-gram not applied yet.
        """
        if self.tokenizer.mask_token is None:
            raise ValueError(
                "This tokenizer does not have a mask token which is necessary for masked language modeling. Remove the --mlm flag if you want to use this tokenizer."
            )

        labels = inputs.clone()
        # We sample a few tokens in each sequence for masked-LM training (with probability args.mlm_probability defaults to 0.15 in Bert/RoBERTa)
        probability_matrix = torch.full(labels.shape, self.mlm_probability)
        special_tokens_mask = [
            self.tokenizer.get_special_tokens_mask(val, already_has_special_tokens=True) for val in labels.tolist()
        ]
        probability_matrix.masked_fill_(torch.tensor(special_tokens_mask, dtype=torch.bool), value=0.0)
        if self.tokenizer._pad_token is not None:
            padding_mask = labels.eq(self.tokenizer.pad_token_id)
            probability_matrix.masked_fill_(padding_mask, value=0.0)
        masked_indices = torch.bernoulli(probability_matrix).bool()
        # probability be `1` (masked), however in albert model attention mask `0` means masked, revert the value
        attention_mask = (~masked_indices).float()
        if self.tokenizer._pad_token is not None:
            attention_padding_mask = labels.eq(self.tokenizer.pad_token_id)
            attention_mask.masked_fill_(attention_padding_mask, value=1.0)
        labels[~masked_indices] = -100  # We only compute loss on masked tokens, -100 is default for CE compute

        # 80% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
        indices_replaced = torch.bernoulli(torch.full(labels.shape, 0.8)).bool() & masked_indices
        inputs[indices_replaced] = self.tokenizer.convert_tokens_to_ids(self.tokenizer.mask_token)

        # 10% of the time, we replace masked input tokens with random word
        indices_random = torch.bernoulli(torch.full(labels.shape, 0.5)).bool() & masked_indices & ~indices_replaced
        random_words = torch.randint(len(self.tokenizer), labels.shape, dtype=torch.long)
        inputs[indices_random] = random_words[indices_random]

        # The rest of the time (10% of the time) we keep the masked input tokens unchanged
        return inputs, labels, attention_mask


@dataclass
class DataCollatorForPermutationLanguageModeling:
    """
    Data collator used for permutation language modeling.

    - collates batches of tensors, honoring their tokenizer's pad_token
    - preprocesses batches for permutation language modeling with procedures specific to XLNet
    """

    tokenizer: PreTrainedTokenizerBase
    plm_probability: float = 1 / 6
    max_span_length: int = 5  # maximum length of a span of masked tokens

    def __call__(
        self, examples: List[Union[List[int], torch.Tensor, Dict[str, torch.Tensor]]]
    ) -> Dict[str, torch.Tensor]:
        if isinstance(examples[0], (dict, BatchEncoding)):
            examples = [e["input_ids"] for e in examples]
        batch = _collate_batch(examples, self.tokenizer)
        inputs, perm_mask, target_mapping, labels = self.mask_tokens(batch)
        return {"input_ids": inputs, "perm_mask": perm_mask, "target_mapping": target_mapping, "labels": labels}

    def mask_tokens(self, inputs: torch.Tensor) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor, torch.Tensor]:
        """
        The masked tokens to be predicted for a particular sequence are determined by the following algorithm:

            0. Start from the beginning of the sequence by setting ``cur_len = 0`` (number of tokens processed so far).
            1. Sample a ``span_length`` from the interval ``[1, max_span_length]`` (length of span of tokens to be
               masked)
            2. Reserve a context of length ``context_length = span_length / plm_probability`` to surround span to be
               masked
            3. Sample a starting point ``start_index`` from the interval ``[cur_len, cur_len + context_length -
               span_length]`` and mask tokens ``start_index:start_index + span_length``
            4. Set ``cur_len = cur_len + context_length``. If ``cur_len < max_len`` (i.e. there are tokens remaining in
               the sequence to be processed), repeat from Step 1.
        """

        if self.tokenizer.mask_token is None:
            raise ValueError(
                "This tokenizer does not have a mask token which is necessary for permutation language modeling. Please add a mask token if you want to use this tokenizer."
            )

        if inputs.size(1) % 2 != 0:
            raise ValueError(
                "This collator requires that sequence lengths be even to create a leakage-free perm_mask. Please see relevant comments in source code for details."
            )

        labels = inputs.clone()
        # Creating the mask and target_mapping tensors
        masked_indices = torch.full(labels.shape, 0, dtype=torch.bool)
        target_mapping = torch.zeros((labels.size(0), labels.size(1), labels.size(1)), dtype=torch.float32)

        for i in range(labels.size(0)):
            # Start from the beginning of the sequence by setting `cur_len = 0` (number of tokens processed so far).
            cur_len = 0
            max_len = labels.size(1)

            while cur_len < max_len:
                # Sample a `span_length` from the interval `[1, max_span_length]` (length of span of tokens to be masked)
                span_length = torch.randint(1, self.max_span_length + 1, (1,)).item()
                # Reserve a context of length `context_length = span_length / plm_probability` to surround the span to be masked
                context_length = int(span_length / self.plm_probability)
                # Sample a starting point `start_index` from the interval `[cur_len, cur_len + context_length - span_length]` and mask tokens `start_index:start_index + span_length`
                start_index = cur_len + torch.randint(context_length - span_length + 1, (1,)).item()
                masked_indices[i, start_index : start_index + span_length] = 1
                # Set `cur_len = cur_len + context_length`
                cur_len += context_length

            # Since we're replacing non-masked tokens with -100 in the labels tensor instead of skipping them altogether,
            # the i-th predict corresponds to the i-th token.
            target_mapping[i] = torch.eye(labels.size(1))

        special_tokens_mask = torch.tensor(
            [self.tokenizer.get_special_tokens_mask(val, already_has_special_tokens=True) for val in labels.tolist()],
            dtype=torch.bool,
        )
        masked_indices.masked_fill_(special_tokens_mask, value=0.0)
        if self.tokenizer._pad_token is not None:
            padding_mask = labels.eq(self.tokenizer.pad_token_id)
            masked_indices.masked_fill_(padding_mask, value=0.0)

        # Mask indicating non-functional tokens, where functional tokens are [SEP], [CLS], padding, etc.
        non_func_mask = ~(padding_mask | special_tokens_mask)

        inputs[masked_indices] = self.tokenizer.mask_token_id
        labels[~masked_indices] = -100  # We only compute loss on masked tokens

        perm_mask = torch.zeros((labels.size(0), labels.size(1), labels.size(1)), dtype=torch.float32)

        for i in range(labels.size(0)):
            # Generate permutation indices i.e. sample a random factorisation order for the sequence. This will
            # determine which tokens a given token can attend to (encoded in `perm_mask`).
            # Note: Length of token sequence being permuted has to be less than or equal to reused sequence length
            # (see documentation for `mems`), otherwise information may leak through due to reuse. In this implementation,
            # we assume that reused length is half of sequence length and permutation length is equal to reused length.
            # This requires that the sequence length be even.

            # Create a linear factorisation order
            perm_index = torch.arange(labels.size(1))
            # Split this into two halves, assuming that half the sequence is reused each time
            perm_index = perm_index.reshape((-1, labels.size(1) // 2)).transpose(0, 1)
            # Permute the two halves such that they do not cross over
            perm_index = perm_index[torch.randperm(labels.size(1) // 2)]
            # Flatten this out into the desired permuted factorisation order
            perm_index = torch.flatten(perm_index.transpose(0, 1))
            # Set the permutation indices of non-masked (non-functional) tokens to the
            # smallest index (-1) so that:
            # (1) They can be seen by all other positions
            # (2) They cannot see masked positions, so there won't be information leak
            perm_index.masked_fill_(~masked_indices[i] & non_func_mask[i], -1)
            # The logic for whether the i-th token can attend on the j-th token based on the factorisation order:
            # 0 (can attend): If perm_index[i] > perm_index[j] or j is neither masked nor a functional token
            # 1 (cannot attend): If perm_index[i] <= perm_index[j] and j is either masked or a functional token
            perm_mask[i] = (
                perm_index.reshape((labels.size(1), 1)) <= perm_index.reshape((1, labels.size(1)))
            ) & masked_indices[i]

        return inputs.long(), perm_mask, target_mapping, labels.long()