carnival13/xnli-cl_trans

这是一个用于文本分类的数据集，包含前提和假设两个文本字段，以及一个标签字段，用于表示两者之间的逻辑关系，包括蕴含(entailment)、中立(neutral)和矛盾(contradiction)。数据集分为训练集、测试集和验证集，共有超过40万条示例。此外，数据集还提供了input_ids、attention_mask和token_type_ids等辅助信息，这些可能是为神经网络模型设计的。

This is a text classification dataset containing two text fields, premise and hypothesis, along with a label field that indicates the logical relationship between them, including entailment, neutral, and contradiction. The dataset is split into training, test, and validation sets, with a total of over 400,000 examples. Additionally, the dataset provides auxiliary information such as input_ids, attention_mask, and token_type_ids, which may be designed for neural network models.