metadata
dataset_info:
features:
- name: paper_arxiv_id
dtype: string
- name: figure_id
dtype: int64
splits:
- name: train
num_bytes: 12364557
num_examples: 546029
download_size: 3509074
dataset_size: 12364557
configs:
- config_name: default
data_files:
- split: train
path: data/train-*