RLHFlow/self_rewarding_turn2_example
收藏官方服务:
资源简介:
该数据集包含了答案、正确答案、是否给予第一次奖励以及提示信息(包括内容和角色)等字段。训练集共有44968个样本,数据集大小为409228915字节。
The dataset includes fields for answers, correct answers, whether the first reward is given, and prompt messages (including content and role). The training set contains 44968 samples, with a total dataset size of 409228915 bytes.
提供机构:
RLHFlow


