README.md

---
dataset_info:
  features:
  - name: document_id
    dtype: string
  - name: page
    dtype: int64
  - name: image_filename
    dtype: string
  - name: image
    dtype: image
  - name: source
    dtype: string
  - name: prompt
    dtype: string
  - name: qa_data
    struct:
    - name: questions
      list:
      - name: answer
        sequence: string
      - name: question
        dtype: string
  - name: query
    dtype: string
  - name: answer
    dtype: string
  - name: model
    dtype: string
  splits:
  - name: train
    num_bytes: 1361609891.305082
    num_examples: 1890
  - name: validation
    num_bytes: 174185044.0
    num_examples: 255
  - name: test
    num_bytes: 325306834.0
    num_examples: 506
  download_size: 1662086187
  dataset_size: 1861101769.305082
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---