metadata
dataset_info:
features:
- name: bibtex_url
dtype: 'null'
- name: proceedings
dtype: string
- name: bibtext
dtype: string
- name: abstract
dtype: string
- name: title
dtype: string
- name: authors
sequence: string
- name: id
dtype: string
- name: type
dtype: string
- name: arxiv_id
dtype: string
- name: GitHub
sequence: string
- name: paper_page
dtype: string
- name: n_linked_authors
dtype: int64
- name: upvotes
dtype: int64
- name: num_comments
dtype: int64
- name: n_authors
dtype: int64
- name: Models
sequence: string
- name: Datasets
sequence: string
- name: Spaces
sequence: string
- name: old_Models
sequence: string
- name: old_Datasets
sequence: string
- name: old_Spaces
sequence: string
- name: paper_page_exists_pre_conf
dtype: int64
splits:
- name: train
num_bytes: 2452147
num_examples: 1148
download_size: 1226457
dataset_size: 2452147
configs:
- config_name: default
data_files:
- split: train
path: data/train-*