HanningZhang/llama3_sft_gsm8k1_star_plus

该数据集名为llama3_sft_gsm8k1_star_plus，包含对话内容、角色、是否为第一轮对话、索引、提示、答案、目标和奖励等字段。对话内容包括对话文本（content）和角色（role），数据类型分别为字符串。是否为第一轮对话（first_round）的数据类型为布尔型，索引（idx）的数据类型为整型。提示（prompt）和答案（answers）的数据类型为字符串，目标（gt）的数据类型为字符串，奖励（rewards）的数据类型为布尔型。数据集分为训练集，共有227103个示例，大小为2,327,254,296字节。

The dataset namedllama3_sft_gsm8k1_star_plus includes fields such as conversation content, role, whether it is the first round of conversation, index, prompt, answer, goal, and reward. The conversation content includes dialogue text (content) and role (role), with data types of string. Whether it is the first round of conversation (first_round) has a data type of boolean, and the index (idx) has a data type of integer. The prompt (prompt) and answer (answers) have data types of string, the goal (gt) has a data type of string, and the reward (rewards) has a data type of boolean. The dataset is divided into a training set, which contains 227,103 examples and is 2,327,254,296 bytes in size.