TAUR-dev/BF16kEval_FinEval_16k_fulleval__3args_ours-eval_rl_letter_countdown_4o

该数据集是一个问答数据集，包含了问题、答案、任务配置、数据源、提示信息以及模型对答案的多种评估和响应信息。数据集还包含了正确率、技能计数和翻转次数等评估指标。数据集分为训练集，共包含1200个示例。

This dataset is a question-answering dataset containing questions, answers, task configurations, data sources, prompt information, and various evaluations and responses from the model. It also includes evaluation metrics such as accuracy, skill count, and flip counts. The dataset is split into a training set with a total of 1200 examples.