---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: answers
    dtype: string
  - name: rationales
    dtype: string
  - name: question_id
    dtype: int64
  splits:
  - name: train
    num_bytes: 42492498.0
    num_examples: 120
  download_size: 22594396
  dataset_size: 42492498.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---