scottgeng00/olmo-3-rlvr-nonreasoner_integration_mix-deltas

这是一个包含prompt、选择(chosen)和拒绝(rejected)内容及其角色的数据集，用于训练和评估模型。数据集分为训练集，共有190000个示例，每个示例包含了prompt文本、选择内容、拒绝内容、选择模型、拒绝模型、数据集来源和真实标签。This dataset includes prompt, chosen and rejected content along with their roles, used for training and evaluating models. The dataset is split into a training set with a total of 190,000 examples, each containing prompt text, chosen content, rejected content, chosen model, rejected model, dataset source, and ground truth label.