shisa-ai/shisa-v2-405b-ultrafeedback-armorm

该数据集包含了与某个提示(prompt)相关的信息，如提示的ID、提示文本、所有生成的响应、每个响应的相关性得分(all_rm_scores)，以及选定的(chosen)和拒绝的(rejected)响应内容及其角色。数据集分为训练集(train)，包含约59569个示例，总大小约为782MB。数据集配置为默认配置，数据文件路径指向训练数据。

The dataset consists of information related to a prompt, including prompt ID, prompt text, all generated responses, relevance scores for each response (all_rm_scores), and the content and roles of chosen and rejected responses. The dataset is split into a training set (train) with approximately 59569 examples, totaling about 782MB in size. The dataset is configured with a default setting, and the data file paths point to the training data.