scottgeng00/rlvr-prompts-mixin_it_up-v2

该数据集包含四个字段：提示(prompt)、数据集名称(dataset)、真实值(ground_truth)和原始HuggingFace源(original_hf_source)，均为文本格式。数据集有一个训练集部分，共有199800个样本，数据集大小为3.68 GB。同时提供了一个默认配置，用于指定训练数据文件的路径。The dataset includes four fields: prompt, dataset name, ground truth, and original HuggingFace source, all in text format. The dataset has a training set with a total of 199800 samples, and the dataset size is 3.68 GB. A default configuration is provided to specify the path to the training data files.