TAUR-dev/D-EVAL__standard_eval_v3__FinEval_16k_fulleval_3args_InstOnly-RL-countdown_5arg-eval_rl

该数据集包含问题、答案以及与任务相关的配置和来源信息。提示内容包含角色和文本信息。数据集还提供了模型响应和相关评估的详细信息，包括正确性评估、最佳答案选择、答案提取和评估的元数据等。数据集分为测试集，共有1000个示例，总体大小为102,205,823字节。

The dataset includes questions, answers, and related configuration and source information for tasks. The prompt content consists of role and text information. The dataset also provides detailed information on model responses and related evaluations, including correctness evaluation, best answer selection, answer extraction, and evaluation metadata. The dataset is split into a test set with a total of 1000 examples, with an overall size of 102,205,823 bytes.