---
dataset_info:
  features:
  - name: image
    struct:
    - name: bytes
      dtype: binary
    - name: path
      dtype: 'null'
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 1123538891
    num_examples: 1000
  download_size: 1122294286
  dataset_size: 1123538891
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---