---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: bboxes
    sequence:
      sequence: int32
  - name: labels
    sequence: int32
  - name: words
    sequence: string
  - name: split
    dtype: string
  splits:
  - name: train
    num_bytes: 1076570989.75
    num_examples: 2890
  download_size: 1065785831
  dataset_size: 1076570989.75
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---