Asap7772/elix_multiuser_eval_4shot_infsft5e7_winrate

该数据集包含了一场对话或相关任务的输入和输出，具体包括对话的提示(prompt)、不同级别的标识符(level_x和level_y，以及对应的ID level_id_x和level_id_y)、不同模型的名称(model_name_x和model_name_y)、模型生成的响应(response_x和response_y)、评分者的级别(scorer_level和对应的ID scorer_level_id)以及标签(label，可能表示正确与否等)。数据集分为训练集(train)，共有1102个样本。

The dataset contains input and output of a conversation or related task, including the conversation prompt, identifiers for different levels (level_x and level_y, along with corresponding IDs level_id_x and level_id_y), names of different models (model_name_x and model_name_y), model-generated responses (response_x and response_y), scorer levels (scorer_level and corresponding IDs scorer_level_id), and labels (label, which may indicate correctness, etc.). The dataset is split into a training set (train) with a total of 1102 samples.