math-extraction-comp/pankajmathur__orca_mini_v9_6_1B-Instruct

该数据集包含以下字段：问题(question)、标准答案(gold)、目标答案(target)、预测答案(prediction)、数据子集(subset)、Qwen提取的答案(qwen_extracted_answer)、Qwen得分(qwen_score)、Harness提取的答案(harness_extracted_answer)、Harness得分(harness_score)、Lighteval-c24870ea提取的答案(lighteval-c24870ea_extracted_answer)和Lighteval-c24870ea得分(lighteval-c24870ea_score)。数据集分为训练集(train)，共有1324个示例，总大小为2,758,051字节。配置信息中包含默认配置(default)，指定了训练集的数据文件路径。

The dataset includes the following fields: question, gold, target, prediction, subset, qwen_extracted_answer, qwen_score, harness_extracted_answer, harness_score, lighteval-c24870ea_extracted_answer, and lighteval-c24870ea_score. The dataset is split into a training set (train) with a total of 1,324 examples and a size of 2,758,051 bytes. The configuration information contains a default configuration that specifies the path to the data files for the training set.