---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: text
    struct:
    - name: English
      dtype: string
    - name: Bengali
      dtype: string
    - name: Hindi
      dtype: string
    - name: Hausa
      dtype: string
    - name: Malayalam
      dtype: string
  splits:
  - name: train
    num_bytes: 1239554697.338
    num_examples: 28927
  - name: validation
    num_bytes: 41678634.0
    num_examples: 998
  download_size: 1278084040
  dataset_size: 1281233331.338
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---