PiC commited on
Commit
396feb9
1 Parent(s): 9d71fea

Change field name "question" to "query"

Browse files
Files changed (1) hide show
  1. phrase_sense_disambiguation.py +4 -4
phrase_sense_disambiguation.py CHANGED
@@ -71,7 +71,7 @@ class PhraseSenseDisambiguation(datasets.GeneratorBasedBuilder):
71
  BUILDER_CONFIGS = [
72
  PSDConfig(
73
  name=_PSD,
74
- version=datasets.Version("1.0.2"),
75
  description="The PiC Dataset for Phrase Sense Disambiguation at short passage level (~22 sentences)"
76
  )
77
  ]
@@ -84,7 +84,7 @@ class PhraseSenseDisambiguation(datasets.GeneratorBasedBuilder):
84
  "id": datasets.Value("string"),
85
  "title": datasets.Value("string"),
86
  "context": datasets.Value("string"),
87
- "question": datasets.Value("string"),
88
  "answers": datasets.Sequence(
89
  {
90
  "text": datasets.Value("string"),
@@ -131,12 +131,12 @@ class PhraseSenseDisambiguation(datasets.GeneratorBasedBuilder):
131
  answer_starts = [answer["answer_start"] for answer in example["answers"]]
132
  answers = [answer["text"] for answer in example["answers"]]
133
 
134
- # Features currently used are "context", "question", and "answers".
135
  # Others are extracted here for the ease of future expansions.
136
  yield key, {
137
  "title": title,
138
  "context": example["context"],
139
- "question": example["question"],
140
  "id": example["id"],
141
  "answers": {
142
  "answer_start": answer_starts,
 
71
  BUILDER_CONFIGS = [
72
  PSDConfig(
73
  name=_PSD,
74
+ version=datasets.Version("1.0.3"),
75
  description="The PiC Dataset for Phrase Sense Disambiguation at short passage level (~22 sentences)"
76
  )
77
  ]
 
84
  "id": datasets.Value("string"),
85
  "title": datasets.Value("string"),
86
  "context": datasets.Value("string"),
87
+ "query": datasets.Value("string"),
88
  "answers": datasets.Sequence(
89
  {
90
  "text": datasets.Value("string"),
 
131
  answer_starts = [answer["answer_start"] for answer in example["answers"]]
132
  answers = [answer["text"] for answer in example["answers"]]
133
 
134
+ # Features currently used are "context", "query", and "answers".
135
  # Others are extracted here for the ease of future expansions.
136
  yield key, {
137
  "title": title,
138
  "context": example["context"],
139
+ "query": example["question"],
140
  "id": example["id"],
141
  "answers": {
142
  "answer_start": answer_starts,