mlfoundations-dev/instruction_filtering_scale_up_math_base_random_filtering_8K

这是一个包含指令种子、来源、推理过程、解决方案、原始行索引、最终推理轨迹以及会话信息的自然语言处理数据集。数据集被划分为一个训练集，共有8000个样本，数据大小为295653139字节。

This is a natural language processing dataset containing instruction seeds, sources, reasoning processes, solutions, original row indices, final reasoning traces, and conversation information. The dataset is split into a training set with a total of 8,000 samples and a data size of 295,653,139 bytes.