purpcode/rl-safety-8k-single-turn
收藏Hugging Face2025-08-10 更新2025-11-01 收录
下载链接：
https://hf-mirror.com/datasets/purpcode/rl-safety-8k-single-turn
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个字段，如数据来源、提示内容、角色、能力、奖励模型以及额外信息。提示内容包含文本内容和角色信息。奖励模型包括地面真实和风格。额外信息则包括数据集名称、索引、预测结果、提示和引用等。数据集分为训练集和测试集，分别包含8192和256个示例。
The dataset includes multiple fields such as data source, prompt content, role, ability, reward model, and extra information. The prompt content includes text and role information. The reward model consists of ground truth and style. Extra information includes dataset name, index, oracles, prompt, and reference. The dataset is split into a training set and a test set, containing 8192 and 256 examples respectively.
提供机构：
purpcode
