---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: annotations
    list:
    - name: caption
      dtype: string
    - name: bbox
      sequence: int64
  splits:
  - name: train
    num_bytes: 1039080370.0
    num_examples: 5000
  download_size: 1046926864
  dataset_size: 1039080370.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---