mygitphase/boolq_indic
收藏Hugging Face2026-05-10 更新2026-05-31 收录
下载链接:
https://hf-mirror.com/datasets/mygitphase/boolq_indic
下载链接
链接失效反馈官方服务:
资源简介:
Indic BoolQ数据集是一个多语言版本的BoolQ(布尔问题)数据集,从英语翻译成10种印度语言。这是一个用于是/否问题的问答数据集,包含约12k自然产生的问题。数据集覆盖的语言包括:孟加拉语(bn)、古吉拉特语(gu)、印地语(hi)、卡纳达语(kn)、马拉地语(mr)、马拉雅拉姆语(ml)、奥里亚语(or)、旁遮普语(pa)、泰米尔语(ta)和泰卢固语(te)。每个示例包含:question(目标语言中的是/否问题)、passage(提供问题背景的段落)、answer(是/否)、label(1表示是,0表示否)和language(ISO 639-1语言代码)。数据集总计约140k个示例,分割大小与原始BoolQ数据集匹配:训练集每种语言9,427个示例,验证集每种语言3,270个示例。
Indic BoolQ Dataset is a multilingual version of the BoolQ (Boolean Questions) dataset, translated from English into 10 Indian languages. It is a question-answering dataset for yes/no questions containing ~12k naturally occurring questions. The dataset includes translations in the following languages: Bengali (bn), Gujarati (gu), Hindi (hi), Kannada (kn), Marathi (mr), Malayalam (ml), Oriya (or), Punjabi (pa), Tamil (ta), and Telugu (te). Each example contains: question (a yes/no question in the target language), passage (a passage providing context for the question), answer (Yes/No), label (1 for yes and 0 for no), and language (ISO 639-1 language code). The total number of examples is ~140k, with split sizes matching the original BoolQ dataset: Training: 9,427 examples per language, Validation: 3,270 examples per language.
提供机构:
mygitphase


