---
dataset_info:
  features:
  - name: token_ids
    list: int64
  - name: token_ids_length
    dtype: int64
  splits:
  - name: train
    num_bytes: 11219883084
    num_examples: 2940889
  - name: valid
    num_bytes: 383951252
    num_examples: 99999
  download_size: 11590380288
  dataset_size: 11603834336
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
---