metadata
dataset_info:
features:
- name: filename
dtype: string
- name: title
dtype: string
- name: abstract
dtype: string
- name: pred_lin
sequence: string
splits:
- name: train
num_bytes: 16668119
num_examples: 13559
download_size: 9848663
dataset_size: 16668119
configs:
- config_name: default
data_files:
- split: train
path: data/train-*