TAUR-dev/D-EVAL__standard_eval_v3__FinEval_16k_fulleval_3args_NoDiv-RL-commonsenseQA-eval_rl

这是一个文本问答数据集，包含问题(question)、答案(answer)、任务配置(task_config)、任务来源(task_source)、提示(prompt)、模型响应(model_responses)等字段。数据集还包含了评估模型的准确性相关的多个字段，如model_responses__eval_is_correct、model_responses__best_of_n_atags__eval_is_correct等。此外，数据集分为测试集(test)。

This is a text question answering dataset that includes fields such as question, answer, task_config, task_source, prompt, model_responses. The dataset also contains multiple fields related to evaluating the accuracy of the model, such as model_responses__eval_is_correct, model_responses__best_of_n_atags__eval_is_correct, etc. In addition, the dataset is split into a test set.