superni-train

该数据集包含15个不同的任务（task002至task1729），每个任务作为一个独立的数据分片。数据集主要包含三个特征字段：'id'（字符串类型，唯一标识符）、'input'（字符串类型，输入内容）和'output'（字符串序列，输出内容）。各任务规模差异显著，最小分片task639包含142个样本（26.9KB），最大分片task511包含1000个样本（2.26MB）。总下载大小5.78MB，解压后数据集总规模9.22MB。数据以分片文件形式存储，路径格式为data/task[编号]-*。This dataset includes 15 distinct tasks, ranging from task002 to task1729, with each task stored as an independent data shard. The dataset primarily contains three feature fields: "id" (string type, unique identifier), "input" (string type, input content), and "output" (string sequence, output content). The scales of these tasks vary greatly: the smallest shard task639 contains 142 samples (26.9 KB), while the largest shard task511 contains 1000 samples (2.26 MB). The total download size of the dataset is 5.78 MB, and the total size of the decompressed dataset is 9.22 MB. The data is stored in the form of shard files, with the path format being data/task[number]-*.