---
dataset_info:
  features:
  - name: token_ids
    sequence: int64
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 176193061000
    num_examples: 10720600
  download_size: 39208740499
  dataset_size: 176193061000
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---