prestonfu/gsm_infinite_hard_r0.5_ops12

该数据集包含5000个训练示例，主要用于问题解决和评估任务。每个示例包括系统提示（system_prompt）、问题（problem）、答案（answer，为浮点数类型）、合成解决方案（synthetic_solution）和评分标准（rubric）。数据集可能涉及数学、科学或逻辑推理等领域，旨在通过系统提示引导生成解决方案，并基于评分标准进行评估。数据以训练集形式提供，总大小为约34.7MB，下载大小约5.2MB。

This dataset contains 5,000 training examples, primarily designed for problem-solving and evaluation tasks. Each example includes a system prompt, a problem, an answer (in float64 format), a synthetic solution, and a rubric. It likely covers domains such as mathematics, science, or logical reasoning, aiming to generate solutions guided by system prompts and evaluate them based on rubrics. The data is provided in a train split, with a total size of approximately 34.7 MB and a download size of about 5.2 MB.