ListOps

该数据集旨在评估模型在列表操作上的性能，其任务属于序列分类。结果显示，FPT模型的性能与全训练的转换器基准测试相媲美。

This dataset is designed to evaluate model performance on list manipulation tasks, which fall under the scope of sequence classification. The results show that the performance of the FPT model is comparable to that of the fully trained Transformer baseline.