README.md

---
dataset_info:
  features:
  - name: question
    dtype: string
  - name: docId
    dtype: int64
  - name: answers
    sequence: string
  - name: data_split
    dtype: string
  - name: bounding_boxes
    sequence:
      sequence: int64
  - name: word_list
    sequence: string
  - name: image_raw
    dtype: image
  - name: ground_truth
    dtype: string
  splits:
  - name: train
    num_bytes: 2266097.0
    num_examples: 10
  - name: valid
    num_bytes: 5902980.0
    num_examples: 10
  - name: test
    num_bytes: 9731023.0
    num_examples: 10
  download_size: 17598517
  dataset_size: 17900100.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
---