math-extraction-comp/T145__ZEUS-8B-V6

该数据集包含问答对和相关评分信息，具体包括问题(question)、标准答案(gold)、目标(target)、预测答案(prediction)、数据子集(subset)、提取的答案(lighteval-4cfbbf17_extracted_answer)以及评分(lighteval-4cfbbf17_score)。数据集目前只有一个训练集(train)，共有1324个样本。

The dataset consists of Q&A pairs along with associated scoring information, including fields for the question, correct answer (gold), target, predicted answer (prediction), subset, extracted answer (lighteval-4cfbbf17_extracted_answer), and score (lighteval-4cfbbf17_score). There is currently one training set (train) with a total of 1324 samples.