---
dataset_info:
  features:
  - name: answer
    dtype: string
  - name: images
    sequence: image
  - name: problem
    dtype: string
  splits:
  - name: train
    num_bytes: 63067838.0
    num_examples: 2000
  - name: test
    num_bytes: 51406891.0
    num_examples: 400
  download_size: 86303813
  dataset_size: 114474729.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---