SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Refactor prepare_seq2seq_batch (#9524)

Browse Source 
* Add target contextmanager and rework prepare_seq2seq_batch

* Fix tests, treat BART and Barthez

* Add last tokenizers

* Fix test

* Set src token before calling the superclass

* Remove special behavior for T5

* Remove needless imports

* Remove needless asserts
This commit is contained in:
Sylvain Gugger
2021-01-12 18:19:38 -05:00
committed by GitHub
parent e6ecef711e
commit 063d8d27f4
24 changed files with 169 additions and 700 deletions
Download Patch File Download Diff File Expand all files Collapse all files

173
src/transformers/tokenization_utils_base.py
View File

@@ -23,6 +23,7 @@ import json
import os
import warnings
from collections import OrderedDict, UserDict
from contextlib import contextmanager
from dataclasses import dataclass, field
from enum import Enum
from typing import TYPE_CHECKING, Any, Dict, List, NamedTuple, Optional, Sequence, Tuple, Union
@@ -1473,68 +1474,6 @@ INIT_TOKENIZER_DOCSTRING = r"""
"""
PREPARE_SEQ2SEQ_BATCH_DOCSTRING = """
Prepare model inputs for translation. For best performance, translate one sentence at a time.
Arguments:
src_texts (:obj:`List[str]`):
List of documents to summarize or source language texts.
tgt_texts (:obj:`list`, `optional`):
List of summaries or target language texts.
max_length (:obj:`int`, `optional`):
Controls the maximum length for encoder inputs (documents to summarize or source language texts) If
left unset or set to :obj:`None`, this will use the predefined model maximum length if a maximum length
is required by one of the truncation/padding parameters. If the model has no specific maximum input
length (like XLNet) truncation/padding to a maximum length will be deactivated.
max_target_length (:obj:`int`, `optional`):
Controls the maximum length of decoder inputs (target language texts or summaries) If left unset or set
to :obj:`None`, this will use the max_length value.
padding (:obj:`bool`, :obj:`str` or :class:`~transformers.tokenization_utils_base.PaddingStrategy`, `optional`, defaults to :obj:`False`):
Activates and controls padding. Accepts the following values:
* :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).
return_tensors (:obj:`str` or :class:`~transformers.tokenization_utils_base.TensorType`, `optional`):
If set, will return tensors instead of list of python integers. Acceptable values are:
* :obj:`'tf'`: Return TensorFlow :obj:`tf.constant` objects.
* :obj:`'pt'`: Return PyTorch :obj:`torch.Tensor` objects.
* :obj:`'np'`: Return Numpy :obj:`np.ndarray` objects.
truncation (:obj:`bool`, :obj:`str` or :class:`~transformers.tokenization_utils_base.TruncationStrategy`, `optional`, defaults to :obj:`True`):
Activates and controls truncation. Accepts the following values:
* :obj:`True` or :obj:`'longest_first'`: Truncate 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. This will truncate token by token, removing a token from the longest sequence in the pair
if a pair of sequences (or a batch of pairs) is provided.
* :obj:`'only_first'`: Truncate 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. This will only
truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
* :obj:`'only_second'`: Truncate 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. This will only
truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
* :obj:`False` or :obj:`'do_not_truncate'` (default): No truncation (i.e., can output batch with
sequence lengths greater than the model maximum admissible input size).
**kwargs:
Additional keyword arguments passed along to :obj:`self.__call__`.
Return:
:class:`~transformers.BatchEncoding`: A :class:`~transformers.BatchEncoding` with the following fields:
- **input_ids** -- List of token ids to be fed to the encoder.
- **attention_mask** -- List of indices specifying which tokens should be attended to by the model.
- **labels** -- List of token ids for tgt_texts.
The full set of keys ``[input_ids, attention_mask, labels]``, will only be returned if tgt_texts is passed.
Otherwise, input_ids, attention_mask will be the only keys.
"""
@add_end_docstrings(INIT_TOKENIZER_DOCSTRING)
class PreTrainedTokenizerBase(SpecialTokensMixin):
"""
@@ -3252,3 +3191,113 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
"indexing errors".format(len(ids), self.model_max_length)
)
self.deprecation_warnings["sequence-length-is-longer-than-the-specified-maximum"] = True
@contextmanager
def as_target_tokenizer(self):
"""
Temporarily sets the tokenizer for encoding the targets. Useful for tokenizer associated to
sequence-to-sequence models that need a slightly different processing for the labels.
"""
yield
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:
"""
Prepare model inputs for translation. For best performance, translate one sentence at a time.
Arguments:
src_texts (:obj:`List[str]`):
List of documents to summarize or source language texts.
tgt_texts (:obj:`list`, `optional`):
List of summaries or target language texts.
max_length (:obj:`int`, `optional`):
Controls the maximum length for encoder inputs (documents to summarize or source language texts) If
left unset or set to :obj:`None`, this will use the predefined model maximum length if a maximum length
is required by one of the truncation/padding parameters. If the model has no specific maximum input
length (like XLNet) truncation/padding to a maximum length will be deactivated.
max_target_length (:obj:`int`, `optional`):
Controls the maximum length of decoder inputs (target language texts or summaries) If left unset or set
to :obj:`None`, this will use the max_length value.
padding (:obj:`bool`, :obj:`str` or :class:`~transformers.tokenization_utils_base.PaddingStrategy`, `optional`, defaults to :obj:`False`):
Activates and controls padding. Accepts the following values:
* :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).
return_tensors (:obj:`str` or :class:`~transformers.tokenization_utils_base.TensorType`, `optional`):
If set, will return tensors instead of list of python integers. Acceptable values are:
* :obj:`'tf'`: Return TensorFlow :obj:`tf.constant` objects.
* :obj:`'pt'`: Return PyTorch :obj:`torch.Tensor` objects.
* :obj:`'np'`: Return Numpy :obj:`np.ndarray` objects.
truncation (:obj:`bool`, :obj:`str` or :class:`~transformers.tokenization_utils_base.TruncationStrategy`, `optional`, defaults to :obj:`True`):
Activates and controls truncation. Accepts the following values:
* :obj:`True` or :obj:`'longest_first'`: Truncate 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. This will truncate token by token, removing a token from the longest sequence in the pair
if a pair of sequences (or a batch of pairs) is provided.
* :obj:`'only_first'`: Truncate 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. This will only
truncate the first sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
* :obj:`'only_second'`: Truncate 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. This will only
truncate the second sequence of a pair if a pair of sequences (or a batch of pairs) is provided.
* :obj:`False` or :obj:`'do_not_truncate'` (default): No truncation (i.e., can output batch with
sequence lengths greater than the model maximum admissible input size).
**kwargs:
Additional keyword arguments passed along to :obj:`self.__call__`.
Return:
:class:`~transformers.BatchEncoding`: A :class:`~transformers.BatchEncoding` with the following fields:
- **input_ids** -- List of token ids to be fed to the encoder.
- **attention_mask** -- List of indices specifying which tokens should be attended to by the model.
- **labels** -- List of token ids for tgt_texts.
The full set of keys ``[input_ids, attention_mask, labels]``, will only be returned if tgt_texts is passed.
Otherwise, input_ids, attention_mask will be the only keys.
"""
# mBART-specific kwargs that should be ignored by other models.
kwargs.pop("src_lang", None)
kwargs.pop("tgt_lang", None)
if max_length is None:
max_length = self.model_max_length
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
with self.as_target_tokenizer():
labels = 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["input_ids"]
return model_inputs