dsrtrain/raft_iter1_sft_new
收藏官方服务：
资源简介：
这是一个包含对话数据的训练集，数据集中的每个样本包括对话内容、角色信息、对话轮次、以及地面真实情况和风格标签。数据集被划分为训练集，共包含7467个示例，总大小约为24268605字节。
This is a training dataset containing conversational data. Each sample in the dataset includes conversation content, role information, conversation turns, as well as ground truth and style labels. The dataset is split into a training set, which contains 7467 examples with a total size of approximately 24268605 bytes.
提供机构：
dsrtrain
