---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: img
    dtype: string
  - name: words
    sequence: string
  - name: bbox
    sequence:
      sequence: int64
  - name: query
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 32768042.528019678
    num_examples: 27077
  - name: test
    num_bytes: 3641431.471980323
    num_examples: 3009
  download_size: 7067160
  dataset_size: 36409474.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---