TAUR-dev/BF16kEval_FinEval_16k_fulleval__3args_rlonly-eval_rl_acronym_5o

该数据集包含了一系列的问题和对应的答案，以及相关的任务配置和来源信息。每个问题可能伴随着一个或多个提示，同时还包括了模型的响应和相关评估指标。数据集还包含了不同任务的技能计数和预算强制延续的详细信息。

The dataset consists of a series of questions and corresponding answers, along with task configurations and source information. Each question may be accompanied by one or more prompts, and includes model responses and related evaluation metrics. The dataset also contains detailed information on skill counts for different tasks and budget-forced continuations.