ttt-ttt9/autocorrelation-spend-time-300s-v13-hyperparams

该数据集是一个包含多个特征字段的训练集，特征包括提示内容(prompt)、数据来源(data_source)、能力(ability)、答案(answer)、原始问题(raw_problem)、难度等级(level)、类型(type)、奖励模型(reward_model)和额外信息(extra_info)。其中，提示内容(prompt)包括内容(content)和角色(role)，奖励模型(reward_model)包括真实标签(ground_truth)和风格(style)，额外信息(extra_info)包括问题ID(problem_id)。数据集仅包含一个训练集split，大小为13210字节。This dataset is a training set containing multiple feature fields, including prompt, data_source, ability, answer, raw_problem, level, type, reward_model, and extra_info. The prompt field includes content and role, the reward_model is a structured field containing ground_truth and style, and the extra_info includes problem_id. The dataset contains only one training split, which is 13210 bytes in size.