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."""
from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

import json
import logging
import os
import regex as re
from io import open

try:
    from functools import lru_cache
except ImportError:
    # Just a dummy decorator to get the checks to run on python2
    # because honestly I don't want to support a byte-level unicode BPE tokenizer on python 2 right now.
    def lru_cache():
        return lambda func: func

from .file_utils import cached_path

logger = logging.getLogger(__name__)

PRETRAINED_VOCAB_ARCHIVE_MAP = {
    'gpt2': "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-vocab.json",
}
PRETRAINED_MERGES_ARCHIVE_MAP = {
    'gpt2': "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-merges.txt",
}
PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP = {
    'gpt2': 1024,
}
VOCAB_NAME = 'vocab.json'
MERGES_NAME = 'merges.txt'

@lru_cache()
def bytes_to_unicode():
    """
    Returns list of utf-8 byte and a corresponding list of unicode strings.
    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 signficant percentage of your normal, say, 32K bpe vocab.
    To avoid that, we want lookup tables between utf-8 bytes and unicode strings.
    And avoids mapping to whitespace/control characters the bpe code barfs on.
    """
    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(object):
    """
    GPT-2 BPE tokenizer. Peculiarities:
        - Byte-level BPE
    """
    @classmethod
    def from_pretrained(cls, pretrained_model_name_or_path, cache_dir=None, *inputs, **kwargs):
        """
        Instantiate a PreTrainedBertModel from a pre-trained model file.
        Download and cache the pre-trained model file if needed.
        """
        if pretrained_model_name_or_path in PRETRAINED_VOCAB_ARCHIVE_MAP:
            vocab_file = PRETRAINED_VOCAB_ARCHIVE_MAP[pretrained_model_name_or_path]
            merges_file = PRETRAINED_MERGES_ARCHIVE_MAP[pretrained_model_name_or_path]
        else:
            vocab_file = os.path.join(pretrained_model_name_or_path, VOCAB_NAME)
            merges_file = os.path.join(pretrained_model_name_or_path, MERGES_NAME)
        # redirect to the cache, if necessary
        try:
            resolved_vocab_file = cached_path(vocab_file, cache_dir=cache_dir)
            resolved_merges_file = cached_path(merges_file, cache_dir=cache_dir)
        except EnvironmentError:
            logger.error(
                "Model name '{}' was not found in model name list ({}). "
                "We assumed '{}' was a path or url but couldn't find files {} and {} "
                "at this path or url.".format(
                    pretrained_model_name_or_path,
                    ', '.join(PRETRAINED_VOCAB_ARCHIVE_MAP.keys()),
                    pretrained_model_name_or_path,
                    vocab_file, merges_file))
            return None
        if resolved_vocab_file == vocab_file and resolved_merges_file == merges_file:
            logger.info("loading vocabulary file {}".format(vocab_file))
            logger.info("loading merges file {}".format(merges_file))
        else:
            logger.info("loading vocabulary file {} from cache at {}".format(
                vocab_file, resolved_vocab_file))
            logger.info("loading merges file {} from cache at {}".format(
                merges_file, resolved_merges_file))
        if pretrained_model_name_or_path in PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP:
            # if we're using a pretrained model, ensure the tokenizer wont index sequences longer
            # than the number of positional embeddings
            max_len = PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP[pretrained_model_name_or_path]
            kwargs['max_len'] = min(kwargs.get('max_len', int(1e12)), max_len)
        # Instantiate tokenizer.
        tokenizer = cls(resolved_vocab_file, resolved_merges_file, *inputs, **kwargs)
        return tokenizer

    def __init__(self, vocab_file, merges_file, errors='replace', max_len=None):
        self.max_len = max_len if max_len is not None else int(1e12)
        self.encoder = json.load(open(vocab_file))
        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+""")

    def __len__(self):
        return len(self.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:
                    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 encode(self, text):
        bpe_tokens = []
        for token in re.findall(self.pat, text):
            token = ''.join(self.byte_encoder[b] for b in token.encode('utf-8'))
            bpe_tokens.extend(self.encoder[bpe_token] for bpe_token in self.bpe(token).split(' '))
        if len(bpe_tokens) > self.max_len:
            logger.warning(
                "Token indices sequence length is longer than the specified maximum "
                " sequence length for this OpenAI GPT-2 model ({} > {}). Running this"
                " sequence through the model will result in indexing errors".format(len(bpe_tokens), self.max_len)
            )
        return bpe_tokens

    def decode(self, tokens):
        text = ''.join([self.decoder[token] for token in tokens])
        text = bytearray([self.byte_decoder[c] for c in text]).decode('utf-8', errors=self.errors)
        return text