metadata
dataset_info:
features:
- name: authorIDs
dtype: string
- name: fullText
dtype: string
- name: documentID
dtype: int64
- name: positive
sequence: int64
- name: cluster
dtype: int64
- name: negative
sequence: int32
splits:
- name: train
num_bytes: 19543170541
num_examples: 3456194
download_size: 17575548944
dataset_size: 19543170541
configs:
- config_name: default
data_files:
- split: train
path: data/train-*