tokenization_gpt2.py

# coding=utf-8
# Copyright 2018 The Open AI Team Authors and The HuggingFace Inc. team.
#
# 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.

"""Tokenization classes for OpenAI GPT (BPE)."""

import json
import logging
import os
from functools import lru_cache
from io import open
from typing import List, Optional

import regex as re

from .tokenization_base import PreTrainedTokenizer

logger = logging.getLogger(__name__)

VOCAB_FILES_NAMES = {
    "vocab_file": "vocab.json",
    "merges_file": "merges.txt",
}

PRETRAINED_VOCAB_FILES_MAP = {
    "vocab_file": {"gpt2": "https://huggingface.co/gpt2/resolve/main/vocab.json"},
    "merges_file": {"gpt2": "https://huggingface.co/gpt2/resolve/main/merges.txt"},
}

PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
    "gpt2": 1024,
}


@lru_cache()
def bytes_to_unicode():
    """
    Returns list of utf-8 byte and a mapping to unicode strings. We specifically avoids mapping
    to whitespace/control characters the bpe code barfs on.

    The reversible bpe codes work on unicode strings. This means you need a large # of unicode
    characters in your vocab if you want to avoid UNKs. When you're at something like a 10B token
    dataset you end up needing around 5K for decent coverage. This is a significant percentage
    of your normal, say, 32K bpe vocab. To avoid that, we want lookup tables between utf-8
    bytes and unicode strings.
    """
    bs = (
        list(range(ord("!"), ord("~") + 1))
        + list(range(ord("¡"), ord("¬") + 1))
        + list(range(ord("®"), ord("ÿ") + 1))
    )
    cs = bs[:]
    n = 0
    for b in range(2 ** 8):
        if b not in bs:
            bs.append(b)
            cs.append(2 ** 8 + n)
            n += 1
    cs = [chr(n) for n in cs]
    return dict(zip(bs, cs))


def get_pairs(word):
    """
    Return set of symbol pairs in a word.

    Word is represented as tuple of symbols (symbols being variable-length strings).
    """
    pairs = set()
    prev_char = word[0]
    for char in word[1:]:
        pairs.add((prev_char, char))
        prev_char = char
    return pairs


class GPT2Tokenizer(PreTrainedTokenizer):
    """
    Construct a GPT-2 tokenizer. Based on byte-level Byte-Pair-Encoding.

    Args:
        vocab_file (:obj:`str`):
            Path to the vocabulary file.
        merges_file (:obj:`str`):
            Path to the merges file.
        errors (:obj:`str`, `optional`, defaults to :obj:`"replace"`):
            Paradigm to follow when decoding bytes to UTF-8. See `bytes.decode
            <https://docs.python.org/3/library/stdtypes.html#bytes.decode>`__ for more information.
        unk_token (:obj:`str`, `optional`, defaults to :obj:`<|endoftext|>`):
            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.
        bos_token (:obj:`str`, `optional`, defaults to :obj:`<|endoftext|>`):
            The beginning of sequence token.
        eos_token (:obj:`str`, `optional`, defaults to :obj:`<|endoftext|>`):
            The end of sequence token.
    """

    vocab_files_names = VOCAB_FILES_NAMES
    pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
    max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES

    def __init__(
        self,
        vocab_file,
        merges_file,
        errors="replace",
        unk_token="<|endoftext|>",
        bos_token="<|endoftext|>",
        eos_token="<|endoftext|>",
        add_bos_token=False,
        **kwargs,
    ):
        super(GPT2Tokenizer, self).__init__(
            bos_token=bos_token, eos_token=eos_token, unk_token=unk_token, **kwargs
        )

        self.encoder = json.load(open(vocab_file, encoding="utf-8"))
        self.decoder = {v: k for k, v in self.encoder.items()}
        self.errors = errors  # how to handle errors in decoding
        self.byte_encoder = bytes_to_unicode()
        self.byte_decoder = {v: k for k, v in self.byte_encoder.items()}
        bpe_data = open(merges_file, encoding="utf-8").read().split("\n")[1:-1]
        bpe_merges = [tuple(merge.split()) for merge in bpe_data]
        self.bpe_ranks = dict(zip(bpe_merges, range(len(bpe_merges))))
        self.cache = {}

        # Should haved added re.IGNORECASE so BPE merges can happen for
        # capitalized versions of contractions
        self.pat = re.compile(
            r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+"""
        )
        self.add_bos_token = add_bos_token

    @property
    def vocab_size(self):
        return len(self.encoder)

    def get_vocab(self):
        return dict(self.encoder, **self.added_tokens_encoder)

    def bpe(self, token):
        if token in self.cache:
            return self.cache[token]
        word = tuple(token)
        pairs = get_pairs(word)

        if not pairs:
            return token

        while True:
            bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float("inf")))
            if bigram not in self.bpe_ranks:
                break
            first, second = bigram
            new_word = []
            i = 0
            while i < len(word):
                try:
                    j = word.index(first, i)
                    new_word.extend(word[i:j])
                    i = j
                except:  # noqa
                    new_word.extend(word[i:])
                    break

                if word[i] == first and i < len(word) - 1 and word[i + 1] == second:
                    new_word.append(first + second)
                    i += 2
                else:
                    new_word.append(word[i])
                    i += 1
            new_word = tuple(new_word)
            word = new_word
            if len(word) == 1:
                break
            else:
                pairs = get_pairs(word)
        word = " ".join(word)
        self.cache[token] = word
        return word

    def _tokenize(self, text):
        """Tokenize a string."""
        bpe_tokens = []
        for token in re.findall(self.pat, text):
            # Maps all our bytes to unicode strings, avoiding control tokens
            # of the BPE (spaces in our case)
            token = "".join(self.byte_encoder[b] for b in token.encode("utf-8"))
            bpe_tokens.extend(bpe_token for bpe_token in self.bpe(token).split(" "))
        return bpe_tokens

    def _convert_token_to_id(self, token):
        """Converts a token (str) in an id using the vocab."""
        return self.encoder.get(token, self.encoder.get(self.unk_token))

    def _convert_id_to_token(self, index):
        """Converts an index (integer) in a token (str) using the vocab."""
        return self.decoder.get(index)

    def convert_tokens_to_string(self, tokens):
        """Converts a sequence of tokens (string) to a single string."""
        text = "".join(tokens)
        text = bytearray([self.byte_decoder[c] for c in text]).decode("utf-8", errors=self.errors)
        return text

    def build_inputs_with_special_tokens(
        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
    ) -> List[int]:
        """Add special tokens to a sequence or a pair of sequence.
        GPT2 format sentence input:

        - single sequence: <|endoftext|> tokens_a
        - pair of sequences: <|endoftext|> tokens_a <|endoftext|> tokens_b

        Args:
            token_ids_0 (List[int]): The token ids of sentence 0.
            token_ids_1 (List[int], optional): The token ids of sentence 1. Defaults to None.

        Returns:
            :obj:`List[str]`: The sequence after adding special toekens.
        """
        if self.add_bos_token:
            bos = [self.bos_token_id]
        else:
            bos = []

        if token_ids_1 is None:
            return bos + token_ids_0

        return bos + token_ids_0 + bos + token_ids_1

    def save_vocabulary(self, save_directory, filename_prefix=None):
        if not os.path.isdir(save_directory):
            logger.error(f"Vocabulary path ({save_directory}) should be a directory")
            return
        vocab_file = os.path.join(
            save_directory,
            (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"],
        )
        merge_file = os.path.join(
            save_directory,
            (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["merges_file"],
        )

        with open(vocab_file, "w", encoding="utf-8") as f:
            f.write(json.dumps(self.encoder, ensure_ascii=False))

        index = 0
        with open(merge_file, "w", encoding="utf-8") as writer:
            writer.write("#version: 0.2\n")
            for bpe_tokens, token_index in sorted(self.bpe_ranks.items(), key=lambda kv: kv[1]):
                if index != token_index:
                    logger.warning(
                        f"Saving vocabulary to {merge_file}: BPE merge indices are not consecutive."
                        " Please check that the tokenizer is not corrupted!"
                    )
                    index = token_index
                writer.write(" ".join(bpe_tokens) + "\n")
                index += 1

        return (vocab_file, merge_file)