---
dataset_info:
  features:
  - name: corrupted
    dtype: string
  - name: original
    dtype: string
  - name: __index_level_0__
    dtype: int64
  splits:
  - name: train
    num_bytes: 875360648
    num_examples: 2000000
  download_size: 375995189
  dataset_size: 875360648
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---