---
dataset_info:
  features:
  - name: image_id
    dtype: string
  - name: image
    dtype: image
  - name: width
    dtype: int64
  - name: height
    dtype: int64
  - name: tokens
    sequence: string
  - name: bboxes
    sequence:
      sequence: float64
  - name: labels
    sequence: string
  - name: target_sequence
    dtype: string
  splits:
  - name: train
    num_bytes: 2636582914.852
    num_examples: 1539
  download_size: 2633866734
  dataset_size: 2636582914.852
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---