---
dataset_info:
  features:
  - name: image
    dtype: string
  - name: relationships_text
    dtype: string
  - name: objects
    sequence: string
  splits:
  - name: train
    num_bytes: 7066
    num_examples: 3
  download_size: 18856
  dataset_size: 7066
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---