---
dataset_info:
  features:
  - name: original
    dtype: string
  - name: corrupted
    dtype: string
  splits:
  - name: train
    num_bytes: 785829
    num_examples: 3123
  download_size: 504928
  dataset_size: 785829
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---