# coding=utf-8 # Copyright 2022 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. import json from pathlib import Path from typing import Dict, List, Tuple import datasets from seacrowd.utils.configs import SEACrowdConfig from seacrowd.utils.constants import Tasks, Licenses, TASK_TO_SCHEMA, SCHEMA_TO_FEATURES _CITATION = """\ @article{,@inproceedings{roy-etal-2020-lareqa, title = "{LAR}e{QA}: Language-Agnostic Answer Retrieval from a Multilingual Pool", author = "Roy, Uma and Constant, Noah and Al-Rfou, Rami and Barua, Aditya and Phillips, Aaron and Yang, Yinfei", editor = "Webber, Bonnie and Cohn, Trevor and He, Yulan and Liu, Yang", booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)", month = nov, year = "2020", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2020.emnlp-main.477", doi = "10.18653/v1/2020.emnlp-main.477", pages = "5919--5930", } """ _DATASETNAME = "xquadr" _DESCRIPTION = """\ XQuAD-R is a retrieval version of the XQuAD dataset (a cross-lingual extractive QA dataset) that is a part of the LAReQA benchmark. Like XQuAD, XQUAD-R is an 11-way parallel dataset, where each question (out of around 1200) appears in 11 different languages and has 11 parallel correct answers across the languages. It is designed so as to include parallel QA pairs across languages, allowing questions to be matched with answers from different languages. The span-tagging task in XQuAD is converted into a retrieval task by breaking up each contextual paragraph into sentences, and treating each sentence as a possible target answer. There are around 1000 candidate answers in each language. """ _HOMEPAGE = "https://github.com/google-research-datasets/lareqa" _LANGUAGES = ["tha", "vie"] _LICENSE = Licenses.CC_BY_SA_4_0.value _LOCAL = False _URLS = { "tha": "https://github.com/google-research-datasets/lareqa/raw/master/xquad-r/th.json", "vie": "https://github.com/google-research-datasets/lareqa/raw/master/xquad-r/vi.json", } _SUPPORTED_TASKS = [Tasks.QUESTION_ANSWERING_RETRIEVAL] _SEACROWD_SCHEMA = f"seacrowd_{TASK_TO_SCHEMA[_SUPPORTED_TASKS[0]].lower()}" # qa _SOURCE_VERSION = "1.1.0" # inside the dataset _SEACROWD_VERSION = "2024.06.20" class XquadRDataset(datasets.GeneratorBasedBuilder): """A retrieval version of the XQuAD dataset (a cross-lingual extractive QA dataset)""" SOURCE_VERSION = datasets.Version(_SOURCE_VERSION) SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION) BUILDER_CONFIGS = [] for subset in _LANGUAGES: BUILDER_CONFIGS += [ SEACrowdConfig( name=f"{_DATASETNAME}_{subset}_source", version=SOURCE_VERSION, description=f"{_DATASETNAME} {subset} source schema", schema="source", subset_id=subset, ), SEACrowdConfig( name=f"{_DATASETNAME}_{subset}_{_SEACROWD_SCHEMA}", version=SEACROWD_VERSION, description=f"{_DATASETNAME} {subset} SEACrowd schema", schema=_SEACROWD_SCHEMA, subset_id=subset, ), ] DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_{_LANGUAGES[0]}_source" def _info(self) -> datasets.DatasetInfo: if self.config.schema == "source": features = datasets.Features( { "paragraphs": datasets.Sequence( { "context": datasets.Value("string"), "qas": datasets.Sequence( { "answers": datasets.Sequence( { "answer_start": datasets.Value("int32"), "text": datasets.Value("string"), } ), "id": datasets.Value("string"), "question": datasets.Value("string"), } ), "sentence_breaks": datasets.Sequence( datasets.Sequence(datasets.Value("int32")) ), "sentences": datasets.Sequence(datasets.Value("string")), } ), "title": datasets.Value("string"), } ) elif self.config.schema == _SEACROWD_SCHEMA: features = SCHEMA_TO_FEATURES[ TASK_TO_SCHEMA[_SUPPORTED_TASKS[0]] ] # qa_features features["meta"] = { "title": datasets.Value("string"), "answers_start": datasets.Sequence(datasets.Value("int32")), "answers_text": datasets.Sequence(datasets.Value("string")), } return datasets.DatasetInfo( description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, license=_LICENSE, citation=_CITATION, ) def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]: """Returns SplitGenerators.""" url = _URLS[self.config.subset_id] data_path = Path(dl_manager.download(url)) return [ datasets.SplitGenerator( name=datasets.Split.TRAIN, gen_kwargs={ "data_path": data_path, }, ), ] def _generate_examples(self, data_path: Path) -> Tuple[int, Dict]: """Yields examples as (key, example) tuples.""" with open(data_path, "r", encoding="utf-8") as file: data = json.load(file) key = 0 for example in data["data"]: if self.config.schema == "source": yield key, example key += 1 elif self.config.schema == _SEACROWD_SCHEMA: for paragraph in example["paragraphs"]: # get sentence breaks (sentences' string stop index) break_list = [breaks[1] for breaks in paragraph["sentence_breaks"]] for qa in paragraph["qas"]: # get answers' string start index answer_starts = [answer["answer_start"] for answer in qa["answers"]] # retrieve answers' relevant sentence answers = [] for start in answer_starts: for i, end in enumerate(break_list): if start < end: answers.append(paragraph["sentences"][i]) break yield key, { "id": str(key), "question_id": qa["id"], # "document_id": None, "question": qa["question"], "type": "retrieval", "choices": [], # escape multiple choice qa seacrowd test "context": paragraph["context"], "answer": answers, "meta": { "title": example["title"], "answers_start": answer_starts, "answers_text": [answer["text"] for answer in qa["answers"]], }, } key += 1