sarvamai/boolq-indic

Indic BoolQ数据集是BoolQ数据集的多语言版本，包含从英语翻译成10种印度语言的布尔问题（是/否问题）。每个示例包含一个问题、提供上下文的段落、答案（是/否）、标签（1表示‘是’，0表示‘否’）和语言代码。数据集分为训练集和验证集，分别包含约9,427和3,270个示例。数据集的使用方法是通过Hugging Face的datasets库加载。

Indic BoolQ is a multilingual version of the BoolQ dataset, containing Boolean questions (yes/no questions) translated from English into 10 Indian languages. Each example includes a question, a passage providing context, an answer (yes/no), a label (1 for yes and 0 for no), and a language code. The dataset is divided into training and validation sets, containing approximately 9,427 and 3,270 examples respectively. The dataset can be loaded using the Hugging Face datasets library.