tokenization_t5.py

# coding=utf-8

""" Tokenization class for model T5. Taken from Huggingface Transformers"""


import os
import re
import warnings
from shutil import copyfile
from typing import Any, Dict, List, Optional, Tuple

import sentencepiece as spm


VOCAB_FILES_NAMES = {"vocab_file": "spiece.model"}

PRETRAINED_VOCAB_FILES_MAP = {
    "vocab_file": {
        "t5-small": "https://huggingface.co/t5-small/resolve/main/spiece.model",
        "t5-base": "https://huggingface.co/t5-base/resolve/main/spiece.model",
        "t5-large": "https://huggingface.co/t5-large/resolve/main/spiece.model",
        "t5-3b": "https://huggingface.co/t5-3b/resolve/main/spiece.model",
        "t5-11b": "https://huggingface.co/t5-11b/resolve/main/spiece.model",
    }
}

PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
    "t5-small": 512,
    "t5-base": 512,
    "t5-large": 512,
    "t5-3b": 512,
    "t5-11b": 512,
}


class EncDecTokenizer():
    """
    Construct a T5 tokenizer. Based on `SentencePiece <https://github.com/google/sentencepiece>`__.

    This tokenizer inherits from :class:`~transformers.PreTrainedTokenizer` which contains most of the main methods.
    Users should refer to this superclass for more information regarding those methods.

    Args:
        vocab_file (:obj:`str`):
            `SentencePiece <https://github.com/google/sentencepiece>`__ file (generally has a `.spm` extension) that
            contains the vocabulary necessary to instantiate a tokenizer.
        eos_token (:obj:`str`, `optional`, defaults to :obj:`"</s>"`):
            The end of sequence token.

            .. note::

                When building a sequence using special tokens, this is not the token that is used for the end of
                sequence. The token used is the :obj:`sep_token`.
        unk_token (:obj:`str`, `optional`, defaults to :obj:`"<unk>"`):
            The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this
            token instead.
        pad_token (:obj:`str`, `optional`, defaults to :obj:`"<pad>"`):
            The token used for padding, for example when batching sequences of different lengths.
        extra_ids (:obj:`int`, `optional`, defaults to 100):
            Add a number of extra ids added to the end of the vocabulary for use as sentinels. These tokens are
            accessible as "<extra_id_{%d}>" where "{%d}" is a number between 0 and extra_ids-1. Extra tokens are
            indexed from the end of the vocabulary up to beginning ("<extra_id_0>" is the last token in the vocabulary
            like in T5 preprocessing see `here
            <https://github.com/google-research/text-to-text-transfer-transformer/blob/9fd7b14a769417be33bc6c850f9598764913c833/t5/data/preprocessors.py#L2117>`__).
        additional_special_tokens (:obj:`List[str]`, `optional`):
            Additional special tokens used by the tokenizer.
        sp_model_kwargs (:obj:`dict`, `optional`):
            Will be passed to the ``SentencePieceProcessor.__init__()`` method. The `Python wrapper for SentencePiece
            <https://github.com/google/sentencepiece/tree/master/python>`__ can be used, among other things, to set:

            - ``enable_sampling``: Enable subword regularization.
            - ``nbest_size``: Sampling parameters for unigram. Invalid for BPE-Dropout.

              - ``nbest_size = {0,1}``: No sampling is performed.
              - ``nbest_size > 1``: samples from the nbest_size results.
              - ``nbest_size < 0``: assuming that nbest_size is infinite and samples from the all hypothesis (lattice)
                using forward-filtering-and-backward-sampling algorithm.

            - ``alpha``: Smoothing parameter for unigram sampling, and dropout probability of merge operations for
              BPE-dropout.

    Attributes:
        sp_model (:obj:`SentencePieceProcessor`):
            The `SentencePiece` processor that is used for every conversion (string, tokens and IDs).
    """

    vocab_files_names = VOCAB_FILES_NAMES
    pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
    max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
    model_input_names = ["input_ids", "attention_mask"]

    def __init__(
        self,
        vocab_file,
        eos_token="</s>",
        unk_token="<unk>",
        pad_token="<pad>",
        extra_ids=100,
        extra_extra_ids=28,
        additional_special_tokens=None,
        sp_model_kwargs: Optional[Dict[str, Any]] = None,
        **kwargs
    ) -> None:
        # Add extra_ids to the special token list
        if extra_ids > 0 and additional_special_tokens is None:
            additional_special_tokens = [f"<extra_id_{i}>" for i in range(extra_ids)]
        elif extra_ids > 0 and additional_special_tokens is not None:
            # Check that we have the right number of extra_id special tokens
            extra_tokens = len(set(filter(lambda x: bool("extra_id" in str(x)), additional_special_tokens)))
            if extra_tokens != extra_ids:
                raise ValueError(
                    f"Both extra_ids ({extra_ids}) and additional_special_tokens ({additional_special_tokens}) are provided to T5Tokenizer. "
                    "In this case the additional_special_tokens must include the extra_ids tokens"
                )

        self.sp_model_kwargs = {} if sp_model_kwargs is None else sp_model_kwargs

        self.eos_token = eos_token
        self.unk_token = unk_token
        self.pad_token = pad_token
        self.extra_ids = extra_ids
        self.additional_sepcial_tokens = additional_special_tokens 

        self.vocab_file = vocab_file
        self._extra_ids = extra_ids
        self._extra_extra_ids = extra_extra_ids

        self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
        self.sp_model.Load(vocab_file)

        self.eos_id = self.convert_token_to_id(self.eos_token)
        self.unk_id = self.convert_token_to_id(self.unk_token)
        self.pad_id = self.convert_token_to_id(self.pad_token)
        
        self.special_tokens = [self.eos_token, self.unk_token, self.pad_token]

    @property
    def real_vocab_size(self):
        return self.sp_model.get_piece_size() + self._extra_ids

    @property
    def vocab_size(self):
        return self.sp_model.get_piece_size() + self._extra_ids + self._extra_extra_ids

    def get_vocab(self):
        vocab = {self.convert_ids_to_tokens(i): i for i in range(self.real_vocab_size)}
        return vocab

    def __getstate__(self):
        state = self.__dict__.copy()
        state["sp_model"] = None
        return state

    def __setstate__(self, d):
        self.__dict__ = d

        # for backward compatibility
        if not hasattr(self, "sp_model_kwargs"):
            self.sp_model_kwargs = {}

        self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
        self.sp_model.Load(self.vocab_file)

    def tokenize(self, text: str) -> List[str]:
        """Take as input a string and return a list of strings (tokens) for words/sub-words"""
        return self.sp_model.encode(text, out_type=str)

    def convert_token_to_id(self, token):
        """Converts a token (str) in an id using the vocab."""
        if token.startswith("<extra_id_"):
            match = re.match(r"<extra_id_(\d+)>", token)
            num = int(match.group(1))
            return self.real_vocab_size - num - 1
        return self.sp_model.piece_to_id(token)

    def convert_id_to_token(self, index):
        """Converts an index (integer) in a token (str) using the vocab."""
        if index < self.sp_model.get_piece_size():
            token = self.sp_model.IdToPiece(index)
        else:
            token = f"<extra_id_{self.real_vocab_size - 1 - index}>"
        return token

    def convert_tokens_to_string(self, tokens):
        """Converts a sequence of tokens (string) in a single string."""
        current_sub_tokens = []
        out_string = ""
        for token in tokens:
            # make sure that special tokens are not decoded using sentencepiece model
            if token in self.additional_sepcial_tokens or token in self.special_tokens:
                out_string += self.sp_model.decode_pieces(current_sub_tokens) + token + " "
                current_sub_tokens = []
            else:
                current_sub_tokens.append(token)
        out_string += self.sp_model.decode_pieces(current_sub_tokens)
        return out_string.strip()

    def convert_tokens_to_ids(self, tokens):
        return [self.convert_token_to_id(x) for x in tokens]

    def convert_ids_to_tokens(self, ids):
        return [self.convert_id_to_token(x) for x in ids]

    def encode(self, text):
        return self.convert_tokens_to_ids(self.tokenize(text))

    def decode(self, ids):
        return self.convert_tokens_to_string(self.convert_ids_to_tokens(ids))

    def get_sentinel_id(self, idx):
        return self.real_vocab_size - idx - 1

    def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
        if not os.path.isdir(save_directory):
            return
        out_vocab_file = os.path.join(
            save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"]
        )

        if os.path.abspath(self.vocab_file) != os.path.abspath(out_vocab_file):
            copyfile(self.vocab_file, out_vocab_file)

        return (out_vocab_file,)