---
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
dataset_info:
  features:
  - name: image
    dtype: image
  - name: ground_truth
    dtype: string
  - name: lang
    dtype: string
  splits:
  - name: train
    num_bytes: 168185040552
    num_examples: 2000000
  - name: validation
    num_bytes: 168041625
    num_examples: 2000
  download_size: 166352491951
  dataset_size: 168353082177
---