sapienzanlp/sciq_italian

SciQ - Italian (IT)数据集是SciQ数据集的意大利语翻译版本，主要用于测试模型在回答需要科学知识的问题时的能力。数据集包含科学相关的问题，每个问题都有一个正确答案和三个干扰项，并且大多数问题都提供了支持段落。数据集包括验证集和测试集，分别包含960行和956行数据。数据集是完全并行的英语和意大利语版本，翻译过程使用了开源工具OBenTO-LLM。

This dataset is an Italian translation of the SciQ dataset, designed for scientific question answering. It includes science-related questions with a correct answer and three distractors, along with a support passage for most questions. The dataset is fully parallel between English and Italian, allowing for comparable evaluations. The translation was done using an open-source LLM tool called OBenTO-LLM, aiming for reproducible and transparent research. The dataset includes validation and test splits with 960 and 956 rows respectively. The format of the dataset includes fields such as id, category, input_text, input_text_translation, choices, choice_translations, label, and metadata.