dawidmt/bigbenchhard

该数据集包含多个子数据集，每个子数据集都包含输入和目标特征，输入和目标都是字符串类型。数据集包含了多个配置名称，包括boolean_expressions, causal_judgement, date_understanding等。每个子数据集都有其特定的数据划分和文件路径。此外，README文件还提供了每个子数据集的训练集大小、字节数和下载大小等信息。

This dataset consists of multiple sub-datasets, each with its own configuration name, features, data splits, and file paths. Each sub-dataset includes input and target features, both of which are strings. The dataset includes various configuration names such as boolean_expressions, causal_judgement, date_understanding, etc. Each sub-dataset has its own specific data splits and file paths. Additionally, the README file provides information on the training set size, byte size, and download size for each sub-dataset.