metadata
dataset_info:
features:
- name: id
dtype: int64
- name: title
dtype: string
- name: text
dtype: string
- name: paragraphs
list:
- name: paragraph_id
dtype: int64
- name: text
dtype: string
- name: title
dtype: string
- name: abstract
dtype: string
- name: date_created
dtype: string
- name: date_modified
dtype: string
- name: templates
sequence: string
- name: url
dtype: string
splits:
- name: train
num_bytes: 41134505564
num_examples: 6879507
download_size: 22646569616
dataset_size: 41134505564
configs:
- config_name: default
data_files:
- split: train
path: data/train-*