---
dataset_info:
  features:
  - name: paper_id
    dtype: string
  - name: questions
    sequence: string
  - name: images
    sequence:
    - name: bytes
      dtype: binary
  splits:
  - name: train
    num_bytes: 56322749127
    num_examples: 5734
  download_size: 55297769554
  dataset_size: 56322749127
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---