ContextualAI/hotpotqa
收藏官方服务：
资源简介：
这是一个基于BEIR基准的HotpotQA数据集的重新排列格式版本，用于多跳推理问答任务。数据集包含字段如_id, query, gt_ids等，并划分为一个测试集。原始数据集来自Stanford NLP的HotpotQA。
This dataset is a reranking-formatted version of the HotpotQA dataset from the BEIR benchmark, used for the task of multi-hop reasoning question answering. The dataset includes fields such as _id, query, gt_ids, and is split into a test set. The original dataset is HotpotQA from Stanford NLP.
提供机构：
ContextualAI
