---
license: mit
dataset_info:
  features:
  - name: words
    sequence: string
  - name: bboxes
    sequence:
      sequence: int64
  - name: ner_tags
    sequence:
      class_label:
        names:
          '0': O
          '1': B-COMPANY
          '2': I-COMPANY
          '3': B-DATE
          '4': I-DATE
          '5': B-ADDRESS
          '6': I-ADDRESS
          '7': B-TOTAL
          '8': I-TOTAL
  - name: images
    dtype: image
  - name: fields
    struct:
    - name: ADDRESS
      dtype: string
    - name: COMPANY
      dtype: string
    - name: DATE
      dtype: string
    - name: TOTAL
      dtype: string
  splits:
  - name: train
    num_bytes: 165837948.0
    num_examples: 626
  - name: test
    num_bytes: 99372961.0
    num_examples: 347
  download_size: 194719908
  dataset_size: 265210909.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---