metadata
dataset_info:
features:
- name: bibcode
dtype: string
- name: abstract
dtype: string
- name: date
dtype: string
- name: identifier
dtype: string
- name: keyword
dtype: string
- name: title
dtype: string
- name: year
dtype: int64
- name: read_count
dtype: float64
- name: cite_read_boost
dtype: float64
- name: esources
dtype: string
- name: citation_count
dtype: float64
- name: arXiv_PDF_Link
dtype: string
- name: splitted_sections
dtype: string
splits:
- name: train
num_bytes: 139049297
num_examples: 1501
download_size: 72073122
dataset_size: 139049297
configs:
- config_name: default
data_files:
- split: train
path: data/train-*