selfcorrexp2/llama31_rr_4_star
收藏Hugging Face2024-12-22 更新2025-04-26 收录
下载链接：
https://hf-mirror.com/datasets/selfcorrexp2/llama31_rr_4_star
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了对话的索引、提示、答案序列、是否为第一轮、正确答案、奖励序列、解决方案序列、标志、轮数和对话内容（包括对话内容和角色）。数据集分为训练集，大小为985,913,837字节，共有58,493个示例。数据集支持默认配置，训练集数据文件路径以data/train-开头。
The dataset includes index, prompt, answer sequence, whether it is the first round, correct answer, reward sequence, solution sequence, flag, turn, and conversation content (including dialogue content and role). The dataset is split into training set, which is 985,913,837 bytes in size and contains 58,493 examples. The dataset supports default configuration, with training set data file paths starting with data/train-.
提供机构：
selfcorrexp2
