---
license: apache-2.0
dataset_info:
  features:
  - name: id
    dtype: string
  - name: raw_image
    dtype: image
  - name: utility_chunk_images
    list: image
  - name: question
    dtype: string
  - name: reference_answer
    dtype: string
  splits:
  - name: train
    num_bytes: 60915679
    num_examples: 49
  download_size: 60778890
  dataset_size: 60915679
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---


The complete dataset will be released after the paper is accepted.

完整的数据集将在论文接收后发布。