cfpark00/toy-multistep-nn_10-na_20-nab_20-seed_2

这是一个包含文本提示和完成文本的数据集，用于训练和测试自然语言处理模型。数据集分为训练集、测试集（包括两个子集），每个集合包含相同数量的示例。它具有四个特征：提示文本(prompts)、完成文本(completions)、被遮蔽的单词数量(num_maskeds)和原始文本(texts)。

This dataset includes text prompts and completion texts for training and testing natural language processing models. The dataset is divided into a training set and two test sets, each containing the same number of examples. It has four features: prompt text (prompts), completion text (completions), number of masked words (num_maskeds), and original text (texts).