zjhhhh/stage1_mean_min_expand_ver2_tokenized_gap_0.22
收藏Hugging Face2025-10-17 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/zjhhhh/stage1_mean_min_expand_ver2_tokenized_gap_0.22
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了用户提示、要求、选择响应、基础响应等文本信息，以及相关的统计数据，如均值和多数。同时，还包括了Qwen提示、选择的文本和令牌数量，以及选择和拒绝的奖励。数据集分为训练集和测试集。
The dataset includes text information such as user prompts, requirements, selection responses, and base responses, as well as related statistical data such as means and majorities. It also includes Qwen prompts, chosen text, token counts, and rewards for selection and rejection. The dataset is split into training and test sets.
提供机构：
zjhhhh
