---
dataset_info:
  features:
  - name: choices
    sequence: string
  - name: view
    dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: rationale
    dtype: string
  - name: pil_image
    dtype: image
  splits:
  - name: train
    num_bytes: 917907502.5
    num_examples: 2500
  download_size: 178347542
  dataset_size: 917907502.5
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---