---
dataset_info:
  features:
  - name: inputs_pretokenized
    dtype: string
  - name: targets_pretokenized
    dtype: string
  splits:
  - name: train
    num_bytes: 1634259567
    num_examples: 1005766
  download_size: 839930738
  dataset_size: 1634259567
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---