reasoningMIA/QWQ_bench_mmlu_pro_distilled

该数据集包含了对话信息，每个样本包括数据集名称、提示、响应、响应索引、答案、正确性、token IDs、token长度、对话（包括对话来源和内容）和会员资格等字段。数据集被划分为训练集，共900个示例，数据集大小为40631632字节。

The dataset contains conversation information, with each sample including fields such as dataset name, prompt, response, response index, answer, correctness, token IDs, token length, conversation (including source and content of the conversation), and membership. The dataset is split into a training set with a total of 900 examples, and the dataset size is 40631632 bytes.