metadata
dataset_info:
features:
- name: abstract
dtype: string
- name: authors
dtype: string
- name: title
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 970272849
num_examples: 1000000
download_size: 572594705
dataset_size: 970272849
configs:
- config_name: default
data_files:
- split: train
path: data/train-*