---
dataset_info:
  features:
  - name: translation
    struct:
    - name: en
      dtype: string
    - name: hi_ng
      dtype: string
    - name: source
      dtype: int64
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 52492917
    num_examples: 189102
  download_size: 23971261
  dataset_size: 52492917
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---