---
dataset_info:
  features:
  - name: token_ids
    sequence: int64
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 268713984
    num_examples: 8192
  download_size: 63800060
  dataset_size: 268713984
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---