metadata
dataset_info:
features:
- name: paper_arxiv_id
dtype: string
- name: category_id
dtype: int64
splits:
- name: train
num_bytes: 1983739
num_examples: 88357
download_size: 499845
dataset_size: 1983739
configs:
- config_name: default
data_files:
- split: train
path: data/train-*