Skywork

该数据集用于训练从多个大型语言模型中衍生出的奖励模型，其中包含了用于训练奖励模型的奖励偏好数据。具体任务是对奖励模型进行训练。This dataset, which contains reward preference data for training reward models derived from multiple large language models (LLMs), is intended for the task of training these reward models.