metadata
dataset_info:
features:
- name: images
list: image
- name: problem
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 6720475481
num_examples: 16152
- name: test
num_bytes: 864387961
num_examples: 2048
download_size: 7583461839
dataset_size: 7584863442
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*