---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  splits:
  - name: train
    num_bytes: 168975915696.0
    num_examples: 20616876
  download_size: 71503236187
  dataset_size: 168975915696.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---