MrPotato commited on
Commit
49b39da
·
1 Parent(s): c89df01

changed tokenization

Browse files
Files changed (1) hide show
  1. ref_seg_ger.py +3 -1
ref_seg_ger.py CHANGED
@@ -57,7 +57,7 @@ _URLS = {
57
 
58
  _LABELS = [
59
  'publisher', 'source', 'url', 'other', 'author', 'editor', 'lpage',
60
- 'volume', 'year', 'issue', 'title', 'fpage', 'identifier', 'edition'
61
  ]
62
 
63
  _FEATURES = datasets.Features(
@@ -229,6 +229,8 @@ class RefSeg(datasets.GeneratorBasedBuilder):
229
 
230
  # tokenized_input = row['token'].split(' ')
231
  tkn = self.TOKENIZER.pre_tokenize_str(row['token'])
 
 
232
  if not tkn:
233
  continue
234
  tokenized_input, offsets = zip(*tkn)
 
57
 
58
  _LABELS = [
59
  'publisher', 'source', 'url', 'other', 'author', 'editor', 'lpage',
60
+ 'volume', 'year', 'issue', 'title', 'fpage', 'edition'
61
  ]
62
 
63
  _FEATURES = datasets.Features(
 
229
 
230
  # tokenized_input = row['token'].split(' ')
231
  tkn = self.TOKENIZER.pre_tokenize_str(row['token'])
232
+ if row['label'] == 'identfier':
233
+ row['label'] = 'other'
234
  if not tkn:
235
  continue
236
  tokenized_input, offsets = zip(*tkn)