---
dataset_info:
  features:
  - name: problem
    dtype: string
  - name: answer
    dtype: string
  - name: source
    dtype: string
  - name: image_url
    dtype: string
  - name: images
    sequence: image
  splits:
  - name: train
    num_bytes: 598064278.25
    num_examples: 2998
  - name: test
    num_bytes: 68099169.0
    num_examples: 300
  download_size: 647927686
  dataset_size: 666163447.25
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---