---
dataset_info:
  features:
  - name: original
    dtype: string
  - name: corrupted
    dtype: string
  splits:
  - name: train
    num_bytes: 172001873.7
    num_examples: 180000
  - name: test
    num_bytes: 19111319.3
    num_examples: 20000
  download_size: 106521879
  dataset_size: 191113193.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---