math-extraction-comp/avemio__GRAG-NEMO-12B-ORPO-HESSIAN-AI

该数据集包含多个字段，其中包括问题(question)、正确答案(gold)、目标(target)、预测(prediction)等字符串字段，以及不同评分系统的得分字段，如lighteval-b200fe81_score、qwen_score和harness_score等浮点数字段。数据集分为训练集(train)，共有1324个示例。但是，README文件中并未提供该数据集的具体描述或其应用场景。

The dataset includes multiple fields such as question, gold, target, prediction as string fields, and score fields for different scoring systems like lighteval-b200fe81_score, qwen_score, and harness_score as float fields. The dataset is split into a training set (train) with a total of 1324 examples. However, the README file does not provide a specific description of the dataset or its use case.