Datasets:
GEM
/

Languages:
Undetermined
Multilinguality:
unknown
Size Categories:
unknown
Language Creators:
unknown
Annotations Creators:
none
Source Datasets:
original
ArXiv:
Tags:
License:
Sebastian Gehrmann commited on
Commit
0ea2525
1 Parent(s): dbccc03
Files changed (1) hide show
  1. xlsum.py +10 -8
xlsum.py CHANGED
@@ -30,11 +30,11 @@ _CITATION = """\
30
 
31
 
32
  _DESCRIPTION = """\
33
- We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally
34
  annotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.
35
  The dataset covers 45 languages ranging from low to high-resource, for many of which no
36
- public dataset is currently available. XL-Sum is highly abstractive, concise,
37
- and of high quality, as indicated by human and intrinsic evaluation.
38
  """
39
 
40
  _HOMEPAGE = "https://github.com/csebuetnlp/xl-sum"
@@ -94,7 +94,7 @@ _LANGUAGES = [
94
 
95
  class Xlsum(datasets.GeneratorBasedBuilder):
96
  VERSION = datasets.Version("2.0.0")
97
-
98
  BUILDER_CONFIGS = [
99
  datasets.BuilderConfig(
100
  name="{}".format(lang),
@@ -111,7 +111,8 @@ class Xlsum(datasets.GeneratorBasedBuilder):
111
  "gem_id": datasets.Value("string"),
112
  "url": datasets.Value("string"),
113
  "title": datasets.Value("string"),
114
- "summary": datasets.Value("string"),
 
115
  "text": datasets.Value("string"),
116
  }
117
  ),
@@ -154,14 +155,15 @@ class Xlsum(datasets.GeneratorBasedBuilder):
154
 
155
  def _generate_examples(self, filepath, split):
156
  """Yields examples as (key, example) tuples."""
157
-
158
  with open(filepath, encoding="utf-8") as f:
159
  for idx_, row in enumerate(f, 1):
160
  data = json.loads(row)
161
  yield idx_, {
162
- "gem_id": f"GEM-xlsum_{self.config.name}-{split}-{idx_}",
163
  "url": data["url"],
164
  "title": data["title"],
165
- "summary": data["summary"],
 
166
  "text": data["text"],
167
  }
 
30
 
31
 
32
  _DESCRIPTION = """\
33
+ We present XLSum, a comprehensive and diverse dataset comprising 1.35 million professionally
34
  annotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics.
35
  The dataset covers 45 languages ranging from low to high-resource, for many of which no
36
+ public dataset is currently available. XL-Sum is highly abstractive, concise,
37
+ and of high quality, as indicated by human and intrinsic evaluation.
38
  """
39
 
40
  _HOMEPAGE = "https://github.com/csebuetnlp/xl-sum"
 
94
 
95
  class Xlsum(datasets.GeneratorBasedBuilder):
96
  VERSION = datasets.Version("2.0.0")
97
+
98
  BUILDER_CONFIGS = [
99
  datasets.BuilderConfig(
100
  name="{}".format(lang),
 
111
  "gem_id": datasets.Value("string"),
112
  "url": datasets.Value("string"),
113
  "title": datasets.Value("string"),
114
+ "target": datasets.Value("string"),
115
+ "references": [datasets.Value("string")],
116
  "text": datasets.Value("string"),
117
  }
118
  ),
 
155
 
156
  def _generate_examples(self, filepath, split):
157
  """Yields examples as (key, example) tuples."""
158
+
159
  with open(filepath, encoding="utf-8") as f:
160
  for idx_, row in enumerate(f, 1):
161
  data = json.loads(row)
162
  yield idx_, {
163
+ "gem_id": f"xlsum_{self.config.name}-{split}-{idx_}",
164
  "url": data["url"],
165
  "title": data["title"],
166
+ "target": data["summary"],
167
+ "references": [data["summary"]],
168
  "text": data["text"],
169
  }