Datasets:

datasets-examples
/

doc-yaml-3

configs:
  - config_name: default
    data_files:
      - split: train
        path: data/*.csv
      - split: test
        path: holdout/*.csv
size_categories:
  - n<1K

[doc] manual configuration 3

This dataset contains two csv files in the data/ directory and one csv file in the holdout/ directory, and a YAML field configs that specifies the data files and splits, using glob expressions.