README.md

metadata

dataset_info:
  features:
    - name: input_ids
      sequence: int32
    - name: attention_mask
      sequence: int8
    - name: labels
      sequence: int64
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 216575980652
      num_examples: 1770133
    - name: test
      num_bytes: 3654598230
      num_examples: 29874
  download_size: 72976886216
  dataset_size: 220230578882
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
task_categories:
  - text2text-generation