sungyub/scicom-context-qa
收藏官方服务：
资源简介：
SciCom Context QA数据集是一个汇集了来自多个高质量问答基准的上下文丰富的科学和常识问答对的合并集合。它结合了4个不同来源的数据，统一为上下文丰富的问答格式，并具有10种不同的模板模式。该数据集包括以下基准：CommonsenseQA、OpenBookQA、QASC和SciQ。所有数据集都已处理为适合训练和评估语言模型的统一输入输出格式。
SciCom Context QA Dataset is a merged collection of context-enriched science and commonsense question-answering pairs from multiple high-quality QA benchmarks. It combines data from 4 different sources into a unified context-enriched question-answer format with 10 diverse template patterns. This dataset includes CommonsenseQA, OpenBookQA, QASC, and SciQ benchmarks. All datasets have been processed into a consistent input-output format suitable for training and evaluating language models.
提供机构：
sungyub
