zjhhhh/game_stage1_largesample_zjhhhh__qwen2.5_3B_Instruct_min_stage1_seed_555134_eta_1e4_step_101
收藏数据链接：
官方服务：
资源简介：
该数据集包含了用于评估模型响应的多个字段，包括提示(prompt)、要求(requirements)、基础响应(base_response)和模型响应(model_response)。此外，还包含了多个评判者(judge)对模型响应的多数投票(majority)和平均(mean)评分。数据集分为训练集(train)，共有500个示例。
The dataset includes multiple fields for evaluating model responses, such as prompt, requirements, base responses (base_response), and model responses (model_response). In addition, it contains the majority vote and average ratings from multiple judges (judge) on the model responses. The dataset is split into a training set (train) with a total of 500 examples.
提供机构：
zjhhhh
