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

Update tokenization_xlm.py

Browse Source
This commit is contained in:
Guillem García Subies
2019-08-20 14:06:17 +02:00
committed by GitHub
parent 933841d903
commit bfd75056b0
1 changed files with 3 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
pytorch_transformers/tokenization_xlm.py
View File

@@ -124,8 +124,9 @@ class XLMTokenizer(PreTrainedTokenizer):
**kwargs) **kwargs)
try: try:
import ftfy import ftfy
import spacy from spacy.lang.en import English
self.nlp = spacy.load('en', disable=['parser', 'tagger', 'ner', 'textcat']) _nlp = English()
self.nlp = nlp.Defaults.create_tokenizer(_nlp)
self.fix_text = ftfy.fix_text self.fix_text = ftfy.fix_text
except ImportError: except ImportError:
logger.warning("ftfy or spacy is not installed using BERT BasicTokenizer instead of SpaCy & ftfy.") logger.warning("ftfy or spacy is not installed using BERT BasicTokenizer instead of SpaCy & ftfy.")