metadata
dataset_info:
features:
- name: image
dtype: image
- name: problem
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 850370062.5
num_examples: 4500
- name: val
num_bytes: 95042725
num_examples: 500
download_size: 638846694
dataset_size: 945412787.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*