mlfoundations-dev/science_verification_scale_up

该数据集包含了问题描述、推理过程、解决方案等字段，适用于训练和评估自然语言处理模型在逻辑推理方面的能力。数据集分为训练集，共有48000个示例。

The dataset includes fields such as problem description, reasoning process, and solution, which are suitable for training and evaluating the logical reasoning capabilities of natural language processing models. The dataset is divided into a training set with a total of 48,000 examples.