marcov/multi_nli_promptsource

该数据集主要用于自然语言推理任务，包含多个特征字段，如promptID、pairID、premise、hypothesis等。数据集分为训练集和验证集（包括匹配和不匹配的验证集），并提供了每个分区的字节大小和样本数量。数据集的标签包括entailment（蕴含）、neutral（中立）和contradiction（矛盾）三类，适用于文本蕴含任务。

This dataset is primarily used for natural language inference tasks, containing multiple feature fields such as promptID, pairID, premise, hypothesis, etc. The dataset is divided into training and validation sets (including matched and mismatched validation sets), with the byte size and number of samples provided for each partition. The datasets labels include entailment, neutral, and contradiction, making it suitable for textual entailment tasks.