rubricreward/llm-metric-ultrafeedback-new

该数据集包含了多个字段，如提示信息(prompt)、来源(source)、输入(input)、模型(model)、分数(score)、详细分数(fine_grained_score)、总体分数(overall_score)、原则(principle)、响应(response)、方面(aspect)和ID标识(id)。数据集分为训练集(train)和测试集(test)，其中训练集包含973248个示例，测试集包含31440个示例。整个数据集的大小为5228383771字节。

The dataset includes multiple fields such as prompt, source, input, model, score, fine_grained_score, overall_score, principle, response, aspect, and id. It is split into a training set (train) with 973248 examples and a test set (test) with 31440 examples. The total size of the dataset is 5228383771 bytes.