mlfoundations-dev/REASONING_evalchemy_32_sharded_gpt-4o-mini

该数据集使用Curator制作，包含数学推理任务的相关数据。数据集中的每个样本都是一个数学问题及其解答。样本数据显示了数学问题的上下文、生成参数、任务名称以及元数据。此外，数据集中的模型输出提供了数学问题的逐步解答。可以通过datasets库中的load_dataset函数加载此数据集。

This dataset, created with Curator, contains data related to mathematical reasoning tasks. Each sample in the dataset consists of a mathematical problem and its solution. The sample data shows the context of the mathematical problem, generation parameters, task name, and metadata. Additionally, the model outputs in the dataset provide step-by-step solutions to the mathematical problems. The dataset can be loaded using the load_dataset function from the datasets library.