dataset_infos.json

{
    "train": {
        "name": "dataset_v1.csv",
        "num_rows": 970,
        "num_columns": 2
    }
}