Datasets:

jeggers
/

big_bench_hard

dataset_info:
  features:
    - name: id
      dtype: string
    - name: input
      dtype: string
    - name: target
      dtype: string
    - name: original_dataset
      dtype: string
  splits:
    - name: train
      num_bytes: 2875717
      num_examples: 6511
  download_size: 637621
  dataset_size: 2875717
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*