---
dataset_info:
  features:
  - name: image
    dtype: string
  - name: annotations
    struct:
    - name: bboxes
      sequence:
        sequence: int64
    - name: ner_tags
      sequence: int64
    - name: path
      dtype: string
    - name: words
      sequence: string
  splits:
  - name: train
    num_bytes: 413533752
    num_examples: 7000
  - name: validation
    num_bytes: 59726150
    num_examples: 1000
  - name: test
    num_bytes: 118725881
    num_examples: 2000
  download_size: 477046778
  dataset_size: 591985783
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---