# coding=utf-8
# Copyright 2018 T5 Authors and 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 class for model T5."""
import os
from shutil import copyfile
from typing import List, Optional, Tuple
from ...file_utils import add_start_docstrings, is_sentencepiece_available
from ...tokenization_utils import BatchEncoding
from ...tokenization_utils_base import PREPARE_SEQ2SEQ_BATCH_DOCSTRING
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import logging
if is_sentencepiece_available():
from .tokenization_t5 import T5Tokenizer
else:
T5Tokenizer = None
logger = logging.get_logger(__name__)
####################################################
# Mapping from the keyword arguments names of Tokenizer `__init__`
# to file names for serializing Tokenizer instances
####################################################
VOCAB_FILES_NAMES = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.json"}
####################################################
# Mapping from the keyword arguments names of Tokenizer `__init__`
# to pretrained vocabulary URL for all the model ids.
####################################################
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",
},
"tokenizer_file": {
"t5-small": "https://huggingface.co/t5-small/resolve/main/tokenizer.json",
"t5-base": "https://huggingface.co/t5-base/resolve/main/tokenizer.json",
"t5-large": "https://huggingface.co/t5-large/resolve/main/tokenizer.json",
"t5-3b": "https://huggingface.co/t5-3b/resolve/main/tokenizer.json",
"t5-11b": "https://huggingface.co/t5-11b/resolve/main/tokenizer.json",
},
}
####################################################
# Mapping from model ids to max length of inputs
####################################################
PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
"t5-small": 512,
"t5-base": 512,
"t5-large": 512,
"t5-3b": 512,
"t5-11b": 512,
}
[docs]class T5TokenizerFast(PreTrainedTokenizerFast):
"""
Construct a "fast" T5 tokenizer (backed by HuggingFace's `tokenizers` library). Based on `Unigram
<https://huggingface.co/docs/tokenizers/python/latest/components.html?highlight=unigram#models>`__.
This tokenizer inherits from :class:`~transformers.PreTrainedTokenizerFast` 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.
"""
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 = ["attention_mask"]
slow_tokenizer_class = T5Tokenizer
prefix_tokens: List[int] = []
def __init__(
self,
vocab_file,
tokenizer_file=None,
eos_token="</s>",
unk_token="<unk>",
pad_token="<pad>",
extra_ids=100,
additional_special_tokens=None,
**kwargs
):
# Add extra_ids to the special token list
if extra_ids > 0 and additional_special_tokens is None:
additional_special_tokens = ["<extra_id_{}>".format(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 special tokens
extra_tokens = len(set(filter(lambda x: bool("extra_id_" in 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"
)
super().__init__(
vocab_file,
tokenizer_file=tokenizer_file,
eos_token=eos_token,
unk_token=unk_token,
pad_token=pad_token,
extra_ids=extra_ids,
additional_special_tokens=additional_special_tokens,
**kwargs,
)
self.vocab_file = vocab_file
self._extra_ids = extra_ids
[docs] def save_vocabulary(self, save_directory: str, filename_prefix: Optional[str] = None) -> Tuple[str]:
if not os.path.isdir(save_directory):
logger.error("Vocabulary path ({}) should be a directory".format(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,)
[docs] def create_token_type_ids_from_sequences(
self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
) -> List[int]:
"""
Create a mask from the two sequences passed to be used in a sequence-pair classification task. T5 does not make
use of token type ids, therefore a list of zeros is returned.
Args:
token_ids_0 (:obj:`List[int]`):
List of IDs.
token_ids_1 (:obj:`List[int]`, `optional`):
Optional second list of IDs for sequence pairs.
Returns:
:obj:`List[int]`: List of zeros.
"""
eos = [self.eos_token_id]
if token_ids_1 is None:
return len(token_ids_0 + eos) * [0]
return len(token_ids_0 + eos + token_ids_1 + eos) * [0]
[docs] @add_start_docstrings(PREPARE_SEQ2SEQ_BATCH_DOCSTRING)
def prepare_seq2seq_batch(
self,
src_texts: List[str],
tgt_texts: Optional[List[str]] = None,
max_length: Optional[int] = None,
max_target_length: Optional[int] = None,
padding: str = "longest",
return_tensors: str = None,
truncation: bool = True,
**kwargs,
) -> BatchEncoding:
if max_length is None:
max_length = self.model_max_length
self.prefix_tokens = []
model_inputs = self(
src_texts,
add_special_tokens=True,
return_tensors=return_tensors,
max_length=max_length,
padding=padding,
truncation=truncation,
**kwargs,
)
if tgt_texts is None:
return model_inputs
# Process tgt_texts
if max_target_length is None:
max_target_length = max_length
# set prefix_tokens for target text
self.prefix_tokens = [self.pad_token_id]
labels_and_decoder_mask = self(
tgt_texts,
add_special_tokens=True,
return_tensors=return_tensors,
padding=padding,
max_length=max_target_length,
truncation=truncation,
**kwargs,
)
model_inputs["labels"] = labels_and_decoder_mask["input_ids"]
self.prefix_tokens = []
return model_inputs