README.md

metadata

size_categories:
  - 100K<n<1M
source_datasets:
  - wikitext
task_categories:
  - text-generation
dataset_info:
  features:
    - name: heading
      dtype: string
    - name: paragraph
      dtype: string
  splits:
    - name: test
      num_bytes: 1381448
      num_examples: 2187
    - name: train
      num_bytes: 574572095
      num_examples: 860879
    - name: validation
      num_bytes: 1219110
      num_examples: 1841
  download_size: 313631586
  dataset_size: 577172653
configs:
  - config_name: default
    data_files:
      - split: test
        path: data/test-*
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*