marcov/tmu_gfm_dataset_promptsource

该数据集包含多个特征，包括源数据、输出、语法、流畅性、意义等，这些特征分别以字符串、整数序列和浮点数的形式存在。数据集主要用于训练目的，包含25326个样本，总大小为26316024.0字节，下载大小为6169234字节。

This dataset includes multiple features such as source, output, grammar, fluency, meaning, etc., which are represented as strings, sequences of integers, and floating-point numbers. The dataset is primarily used for training, containing 25,326 samples with a total size of 26,316,024.0 bytes and a download size of 6,169,234 bytes.