---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: objects
    struct:
    - name: bbox
      sequence:
        sequence: float32
    - name: categories
      sequence: string
  splits:
  - name: train
    num_bytes: 136572985.0
    num_examples: 823
  download_size: 121463771
  dataset_size: 136572985.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---