Open Mind Common Sense (OMCS) corpus

该数据集名为OMCS，是一个众包的常识语句知识数据库，包含了来自超过15000名贡献者的百万条句子。该OMCS数据集被用于在CosmosQA数据集上进行微调之前，对文本证据生成器进行预训练。其规模达到百万条句子，任务是对文本证据生成器进行预训练。

This dataset, named OMCS, is a crowdsourced commonsense sentence knowledge database containing millions of sentences from over 15,000 contributors. The OMCS dataset is utilized to pre-train textual evidence generators prior to their fine-tuning on the CosmosQA dataset. With a corpus of millions of sentences, its core task is to pre-train textual evidence generators.