cjziems/Psych-101-Scaling

这是一个包含多个数据集配置的集合，每个配置都有详细的特征、数据类型和分割信息。特征包括文本、目标、近似上下文长度、实验、参与者、论文、选项和一个索引级别。每个数据集还包括测试分割的示例数量、字节和下载大小信息。数据集以与之相关的论文命名，并为每个数据集提供了数据文件路径。

This is a collection of multiple dataset configurations, each with detailed information on features, data types, and splits. The features include text, target, approx_context_length, experiment, participant, paper, options, and an index level. Each dataset also includes information on the number of examples, bytes, and download size for the test split. The datasets are named after the papers they are associated with, and the data files paths are provided for each dataset.