rubricreward/R3-Random-20K
收藏Hugging Face2025-07-30 更新2025-08-09 收录
下载链接：
https://hf-mirror.com/datasets/rubricreward/R3-Random-20K
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ID、原始ID、来源、提示、实际得分、推理轨迹、响应、评分和令牌长度等。这些字段表明数据集可能用于评估某种文本生成或文本处理任务，例如文本摘要或问答系统。训练集包含20000个示例，数据集的总大小为201514456字节。数据集的配置信息显示，训练数据存储在特定的路径下。
The dataset includes multiple fields such as ID, original ID, source, prompt, actual score, reasoning trace, response, score, and token length. These fields suggest that the dataset might be used for evaluating a text generation or text processing task, such as text summarization or a question-answering system. The training set contains 20,000 examples, and the total size of the dataset is 201,514,456 bytes. The configuration information of the dataset shows that the training data is stored under specific paths.
提供机构：
rubricreward
