AtAndDev/simpo-sedir-marco-o1

该数据集名为simpo-sedir-marco-o1，包含1000个训练样本，每个样本包含prompt、all_generated_responses、all_rm_scores、chosen和rejected等特征。prompt和chosen、rejected是字符串类型，all_generated_responses是字符串序列，all_rm_scores是浮点数序列。数据集总大小为26915409字节，下载大小为10435842字节。

The dataset named simpo-sedir-marco-o1 contains 1000 training samples, each with features including prompt, all_generated_responses, all_rm_scores, chosen, and rejected. The prompt, chosen, and rejected are of string type, all_generated_responses is a sequence of strings, and all_rm_scores is a sequence of float64. The total size of the dataset is 26915409 bytes, and the download size is 10435842 bytes.