Datasets:
Tasks:
Question Answering
Modalities:
Text
Formats:
parquet
Sub-tasks:
multiple-choice-qa
Size:
100K - 1M
ArXiv:
License:
File size: 10,812 Bytes
449984e f315d37 449984e f315d37 449984e f315d37 449984e f315d37 449984e f315d37 449984e f315d37 449984e f315d37 449984e f315d37 977bf5b f315d37 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 |
# coding=utf-8
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""EXAMS: a benchmark dataset for multilingual and cross-lingual question answering"""
import json
import datasets
_CITATION = """\
@article{hardalov2020exams,
title={EXAMS: A Multi-subject High School Examinations Dataset for Cross-lingual and Multilingual Question Answering},
author={Hardalov, Momchil and Mihaylov, Todor and Dimitrina Zlatkova and Yoan Dinkov and Ivan Koychev and Preslav Nvakov},
journal={arXiv preprint arXiv:2011.03080},
year={2020}
}
"""
_DESCRIPTION = """\
EXAMS is a benchmark dataset for multilingual and cross-lingual question answering from high school examinations.
It consists of more than 24,000 high-quality high school exam questions in 16 languages,
covering 8 language families and 24 school subjects from Natural Sciences and Social Sciences, among others.
"""
_HOMEPAGE = "https://github.com/mhardalov/exams-qa"
_LICENSE = "CC-BY-SA-4.0"
_URLS_LIST = [
("alignments", "https://github.com/mhardalov/exams-qa/raw/main/data/exams/parallel_questions.jsonl"),
]
_URLS_LIST += [
(
"multilingual_train",
"https://github.com/mhardalov/exams-qa/raw/main/data/exams/multilingual/train.jsonl.tar.gz",
),
("multilingual_dev", "https://github.com/mhardalov/exams-qa/raw/main/data/exams/multilingual/dev.jsonl.tar.gz"),
("multilingual_test", "https://github.com/mhardalov/exams-qa/raw/main/data/exams/multilingual/test.jsonl.tar.gz"),
(
"multilingual_with_para_train",
"https://github.com/mhardalov/exams-qa/raw/main/data/exams/multilingual/with_paragraphs/train_with_para.jsonl.tar.gz",
),
(
"multilingual_with_para_dev",
"https://github.com/mhardalov/exams-qa/raw/main/data/exams/multilingual/with_paragraphs/dev_with_para.jsonl.tar.gz",
),
(
"multilingual_with_para_test",
"https://github.com/mhardalov/exams-qa/raw/main/data/exams/multilingual/with_paragraphs/test_with_para.jsonl.tar.gz",
),
]
_CROSS_LANGUAGES = ["bg", "hr", "hu", "it", "mk", "pl", "pt", "sq", "sr", "tr", "vi"]
_URLS_LIST += [
("crosslingual_test", "https://github.com/mhardalov/exams-qa/raw/main/data/exams/cross-lingual/test.jsonl.tar.gz"),
(
"crosslingual_with_para_test",
"https://github.com/mhardalov/exams-qa/raw/main/data/exams/cross-lingual/with_paragraphs/test_with_para.jsonl.tar.gz",
),
]
for ln in _CROSS_LANGUAGES:
_URLS_LIST += [
(
f"crosslingual_{ln}_train",
f"https://github.com/mhardalov/exams-qa/raw/main/data/exams/cross-lingual/train_{ln}.jsonl.tar.gz",
),
(
f"crosslingual_with_para_{ln}_train",
f"https://github.com/mhardalov/exams-qa/raw/main/data/exams/cross-lingual/with_paragraphs/train_{ln}_with_para.jsonl.tar.gz",
),
(
f"crosslingual_{ln}_dev",
f"https://github.com/mhardalov/exams-qa/raw/main/data/exams/cross-lingual/dev_{ln}.jsonl.tar.gz",
),
(
f"crosslingual_with_para_{ln}_dev",
f"https://github.com/mhardalov/exams-qa/raw/main/data/exams/cross-lingual/with_paragraphs/dev_{ln}_with_para.jsonl.tar.gz",
),
]
_URLs = dict(_URLS_LIST)
class ExamsConfig(datasets.BuilderConfig):
def __init__(self, lang, with_para, **kwargs):
super(ExamsConfig, self).__init__(version=datasets.Version("1.0.0", ""), **kwargs)
self.lang = lang
self.with_para = "_with_para" if with_para else ""
class Exams(datasets.GeneratorBasedBuilder):
"""Exams dataset"""
VERSION = datasets.Version("1.0.0")
BUILDER_CONFIG_CLASS = ExamsConfig
BUILDER_CONFIGS = [
ExamsConfig(
lang="",
with_para=False,
name="alignments",
description="loads the alignment between question IDs across languages",
),
ExamsConfig(
lang="all",
with_para=False,
name="multilingual",
description="Loads the unified multilingual train/dev/test split",
),
ExamsConfig(
lang="all",
with_para=True,
name="multilingual_with_para",
description="Loads the unified multilingual train/dev/test split with Wikipedia support paragraphs",
),
ExamsConfig(
lang="all", with_para=False, name="crosslingual_test", description="Loads crosslingual test set only"
),
ExamsConfig(
lang="all",
with_para=True,
name="crosslingual_with_para_test",
description="Loads crosslingual test set only with Wikipedia support paragraphs",
),
]
for ln in _CROSS_LANGUAGES:
BUILDER_CONFIGS += [
ExamsConfig(
lang=ln,
with_para=False,
name=f"crosslingual_{ln}",
description=f"Loads crosslingual train and dev set for {ln}",
),
ExamsConfig(
lang=ln,
with_para=True,
name=f"crosslingual_with_para_{ln}",
description=f"Loads crosslingual train and dev set for {ln} with Wikipedia support paragraphs",
),
]
DEFAULT_CONFIG_NAME = (
"multilingual_with_para" # It's not mandatory to have a default configuration. Just use one if it make sense.
)
def _info(self):
if self.config.name == "alignments": # This is the name of the configuration selected in BUILDER_CONFIGS above
features = datasets.Features(
{
"source_id": datasets.Value("string"),
"target_id_list": datasets.Sequence(datasets.Value("string")),
}
)
else: # This is an example to show how to have different features for "first_domain" and "second_domain"
features = datasets.Features(
{
"id": datasets.Value("string"),
"question": {
"stem": datasets.Value("string"),
"choices": datasets.Sequence(
{
"text": datasets.Value("string"),
"label": datasets.Value("string"),
"para": datasets.Value("string"),
}
),
},
"answerKey": datasets.Value("string"),
"info": {
"grade": datasets.Value("int32"),
"subject": datasets.Value("string"),
"language": datasets.Value("string"),
},
}
)
return datasets.DatasetInfo(
description=_DESCRIPTION,
features=features, # Here we define them above because they are different between the two configurations
supervised_keys=None,
homepage=_HOMEPAGE,
license=_LICENSE,
citation=_CITATION,
)
def _split_generators(self, dl_manager):
"""Returns SplitGenerators."""
archives = dl_manager.download(_URLs)
if self.config.name == "alignments":
return [
datasets.SplitGenerator(
name="full",
gen_kwargs={"filepath": archives["alignments"]},
),
]
elif self.config.name in ["multilingual", "multilingual_with_para"]:
return [
datasets.SplitGenerator(
name=spl_enum,
gen_kwargs={
"filepath": f"{spl}{self.config.with_para}.jsonl",
"files": dl_manager.iter_archive(archives[f"{self.config.name}_{spl}"]),
},
)
for spl, spl_enum in [
("train", datasets.Split.TRAIN),
("dev", datasets.Split.VALIDATION),
("test", datasets.Split.TEST),
]
]
elif self.config.name in ["crosslingual_test", "crosslingual_with_para_test"]:
return [
datasets.SplitGenerator(
name=datasets.Split.TEST,
gen_kwargs={
"filepath": f"test{self.config.with_para}.jsonl",
"files": dl_manager.iter_archive(archives[self.config.name]),
},
),
]
else:
return [
datasets.SplitGenerator(
name=spl_enum,
gen_kwargs={
"filepath": f"{spl}_{self.config.lang}{self.config.with_para}.jsonl",
"files": dl_manager.iter_archive(archives[f"{self.config.name}_{spl}"]),
},
)
for spl, spl_enum in [
("train", datasets.Split.TRAIN),
("dev", datasets.Split.VALIDATION),
]
]
def _generate_examples(self, filepath, files=None):
if self.config.name == "alignments":
with open(filepath, encoding="utf-8") as f:
for id_, line in enumerate(f):
line_dict = json.loads(line.strip())
in_id, out_list = list(line_dict.items())[0]
yield id_, {"source_id": in_id, "target_id_list": out_list}
else:
for path, f in files:
if path == filepath:
for id_, line in enumerate(f):
line_dict = json.loads(line.strip())
for choice in line_dict["question"]["choices"]:
choice["para"] = choice.get("para", "")
yield id_, {
"id": line_dict["id"],
"question": line_dict["question"],
"answerKey": line_dict["answerKey"],
"info": line_dict["info"],
}
break
|