RyanYr/self-reflect_mini8Bit-t0_sft-t1_om2-1_2_binlabel

该数据集包含80,000个训练样本，每个样本包含问题、生成的解决方案、答案、问题来源以及多个响应和其正确性。数据集的特征字段包括problem（问题）、generated_solution（生成的解决方案）、answer（答案）、problem_source（问题来源）、response@0（响应0）、response@1（响应1）、response@0_ans（响应0的答案）和response@0_correctness（响应0的正确性）。数据集的总大小为821,034,294字节，下载大小为303,685,266字节。

This dataset is primarily used for training models to solve specific problems, including descriptions of problems, generated solutions, correct answers, and sources of problems. The dataset also includes multiple response features, which may be used to evaluate the correctness of different solutions. The dataset is divided into a training set with 80000 samples, with a total size of 821034294 bytes and a download size of 303685266 bytes.