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

Doc + remove artefacts

Browse Source
This commit is contained in:
LysandreJik
2019-09-19 09:42:13 +02:00
parent 6393261e41
commit 8cba057260
1 changed files with 2 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files

41
pytorch_transformers/tokenization_utils.py
View File

@@ -724,9 +724,8 @@ class PreTrainedTokenizer(object):
def encode_plus(self, text, text_pair=None, add_special_tokens=False, output_mask=False, max_length=None, **kwargs):
"""
Converts a string in a sequence of ids (integer), using the tokenizer and vocabulary.
Same as doing ``self.convert_tokens_to_ids(self.tokenize(text))``.
Returns a dictionary containing the encoded sequence or sequence pair. Other values can be returned by this
method: the mask for sequence classification and the overflowing elements if a ``max_length`` is specified.
Args:
text: The first sequence to be encoded.
@@ -801,42 +800,6 @@ class PreTrainedTokenizer(object):
return information
if text_pair is None:
if add_special_tokens:
sequence_tokens = self.convert_tokens_to_ids(self.tokenize(text, **kwargs))
if max_length:
sequence_tokens = sequence_tokens[:max_length - self.num_added_tokens()]
return self.add_special_tokens_single_sentence(sequence_tokens)
else:
ids = self.convert_tokens_to_ids(self.tokenize(text, **kwargs))
return ids[:max_length] if max_length != -1 else ids
first_sentence_tokens = [self._convert_token_to_id(token) for token in self.tokenize(text, **kwargs)]
second_sentence_tokens = [self._convert_token_to_id(token) for token in self.tokenize(text_pair, **kwargs)]
if add_special_tokens:
if max_length:
if len(first_sentence_tokens) + self.num_added_tokens(pair=True) >= max_length:
logger.warning(
"The first sequence is longer than the maximum specified length. This sequence will not be truncated.")
else:
if len(second_sentence_tokens) + len(first_sentence_tokens) + self.num_added_tokens(
pair=True) > max_length:
second_sentence_tokens = second_sentence_tokens[
:max_length - len(first_sentence_tokens) - self.num_added_tokens(
pair=True)]
return self.add_special_tokens_sentences_pair(first_sentence_tokens, second_sentence_tokens,
output_mask)
else:
if max_length:
first_sentence_tokens = first_sentence_tokens[:max_length]
second_sentence_tokens = second_sentence_tokens[:max_length]
if output_mask:
logger.warning("Can't output mask if you're not joining two sequences.")
return first_sentence_tokens, second_sentence_tokens
def add_special_tokens_single_sentence(self, token_ids):
logger.warning("This tokenizer does not make use of special tokens. The sequence has been returned with no modification.")
return token_ids