math-extraction-comp/saltlux__luxia-21.4b-alignment-v1.0

该数据集是一个文本数据集，用于某种问答系统或文本匹配任务。它包含了问题、标准答案（gold）、模型预测的答案（target）、预测答案（prediction）、数据子集标识（subset）、以及由不同工具（harness、qwen、lighteval-0f21c935）提取的答案和对应的评分。数据集分为训练集，共有1324个示例，总大小为2,750,890字节。

This dataset is a text dataset designed for some kind of question-answering or text matching task. It includes questions, reference answers (gold), predicted answers (target), predicted answers (prediction), subset identification, and answers extracted by different tools (harness, qwen, lighteval-0f21c935) along with corresponding scores. The dataset is split into a training set, which contains 1324 examples and has a total size of 2,750,890 bytes.