---
dataset_info:
  features:
  - name: tokens
    sequence: uint16
  splits:
  - name: train
    num_bytes: 19711901592
    num_examples: 985398
  - name: validation
    num_bytes: 200040000
    num_examples: 10000
  download_size: 17632151409
  dataset_size: 19911941592
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---