ASSIN2

该数据集旨在识别葡萄牙语中的文本蕴含关系，模型在此次任务中取得了新的领先成绩，相较于之前的记录，以0.61的F1宏观分数点超越了旧有水平。其规模根据模型的大小和配置而有所不同，所涉及的任务是文本蕴含。

This dataset is designed to recognize textual entailment in Portuguese. Models trained on this dataset have achieved a new state-of-the-art performance on this task, surpassing the previous record by 0.61 macro-F1 score points. The scale of the dataset varies depending on the model size and configuration, and the involved task is textual entailment.