---
dataset_info:
  features:
  - name: images
    list: image
  - name: question
    dtype: string
  - name: choices
    list: string
  - name: answer_idx
    dtype: int32
  - name: datatype
    dtype: string
  splits:
  - name: train
    num_bytes: 395453728
    num_examples: 4500
  download_size: 377337009
  dataset_size: 395453728
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---