baber commited on
Commit
ccfc255
1 Parent(s): b63e244

added description

Browse files
Files changed (1) hide show
  1. agieval.py +11 -13
agieval.py CHANGED
@@ -31,10 +31,6 @@ _CITATION = """\
31
  doi={10.1109/TASLP.2023.3293046}}
32
  """
33
 
34
- _DESCRIPTION = """\
35
- The dataset is an amendment and re-annotation of LogiQA in 2020, a large-scale logical reasoning reading comprehension dataset adapted from the Chinese Civil Service Examination. We increase the data size, refine the texts with manual translation by professionals, and improve the quality by removing items with distinctive cultural features like Chinese idioms. Furthermore, we conduct a fine-grained annotation on the dataset and turn it into a two-way natural language inference (NLI) task, resulting in 35k premise-hypothesis pairs with gold labels, making it the first large-scale NLI dataset for complex logical reasoning
36
- """
37
-
38
  _HOMEPAGE = "https://github.com/csitfun/LogiQA2.0/tree/main"
39
 
40
  _LICENSE = (
@@ -42,6 +38,8 @@ _LICENSE = (
42
  )
43
  HEAD = 'https://raw.githubusercontent.com/microsoft/AGIEval/main/data/v1/'
44
 
 
 
45
  _URLS = {
46
  "sat_en": {
47
  "test": HEAD + 'sat-en.jsonl',
@@ -86,47 +84,47 @@ class AgiEval(datasets.GeneratorBasedBuilder):
86
  datasets.BuilderConfig(
87
  name="aqua_rat",
88
  version=VERSION,
89
- description="",
90
  ),
91
  datasets.BuilderConfig(
92
  name="sat_en",
93
  version=VERSION,
94
- description="",
95
  ),
96
  datasets.BuilderConfig(
97
  name="sat_en_wop",
98
  version=VERSION,
99
- description="",
100
  ),
101
  datasets.BuilderConfig(
102
  name="sat_math",
103
  version=VERSION,
104
- description="",
105
  ),
106
  datasets.BuilderConfig(
107
  name="lsat_ar",
108
  version=VERSION,
109
- description="",
110
  ),
111
  datasets.BuilderConfig(
112
  name="lsat_lr",
113
  version=VERSION,
114
- description="",
115
  ),
116
  datasets.BuilderConfig(
117
  name="lsat_rc",
118
  version=VERSION,
119
- description="",
120
  ),
121
  datasets.BuilderConfig(
122
  name="logiqa",
123
  version=VERSION,
124
- description="",
125
  ),
126
  datasets.BuilderConfig(
127
  name="math_agieval",
128
  version=VERSION,
129
- description="",
130
  ),
131
  ]
132
  DEFAULT_CONFIG_NAME = "aqua_rat"
 
31
  doi={10.1109/TASLP.2023.3293046}}
32
  """
33
 
 
 
 
 
34
  _HOMEPAGE = "https://github.com/csitfun/LogiQA2.0/tree/main"
35
 
36
  _LICENSE = (
 
38
  )
39
  HEAD = 'https://raw.githubusercontent.com/microsoft/AGIEval/main/data/v1/'
40
 
41
+ _DESCRIPTION = "AGIEval is a human-centric benchmark specifically designed to evaluate the general abilities of foundation models in tasks pertinent to human cognition and problem-solving. This benchmark is derived from 20 official, public, and high-standard admission and qualification exams intended for general human test-takers, such as general college admission tests"
42
+
43
  _URLS = {
44
  "sat_en": {
45
  "test": HEAD + 'sat-en.jsonl',
 
84
  datasets.BuilderConfig(
85
  name="aqua_rat",
86
  version=VERSION,
87
+ description=_DESCRIPTION,
88
  ),
89
  datasets.BuilderConfig(
90
  name="sat_en",
91
  version=VERSION,
92
+ description=_DESCRIPTION,
93
  ),
94
  datasets.BuilderConfig(
95
  name="sat_en_wop",
96
  version=VERSION,
97
+ description=_DESCRIPTION,
98
  ),
99
  datasets.BuilderConfig(
100
  name="sat_math",
101
  version=VERSION,
102
+ description=_DESCRIPTION,
103
  ),
104
  datasets.BuilderConfig(
105
  name="lsat_ar",
106
  version=VERSION,
107
+ description=_DESCRIPTION,
108
  ),
109
  datasets.BuilderConfig(
110
  name="lsat_lr",
111
  version=VERSION,
112
+ description=_DESCRIPTION,
113
  ),
114
  datasets.BuilderConfig(
115
  name="lsat_rc",
116
  version=VERSION,
117
+ description=_DESCRIPTION,
118
  ),
119
  datasets.BuilderConfig(
120
  name="logiqa",
121
  version=VERSION,
122
+ description=_DESCRIPTION,
123
  ),
124
  datasets.BuilderConfig(
125
  name="math_agieval",
126
  version=VERSION,
127
+ description=_DESCRIPTION,
128
  ),
129
  ]
130
  DEFAULT_CONFIG_NAME = "aqua_rat"