---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 6278459538.220414
    num_examples: 20045
  - name: test
    num_bytes: 322990025.8255863
    num_examples: 1056
  download_size: 6470762421
  dataset_size: 6601449564.0460005
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---