metadata
license: mit
dataset_info:
features:
- name: ads_id
dtype: string
- name: arxiv_id
dtype: string
- name: title
dtype: string
- name: abstract
dtype: string
- name: embed
sequence: float32
- name: umap_x
dtype: float32
- name: umap_y
dtype: float32
- name: date
dtype: date32
- name: cites
dtype: int64
- name: bibcode
dtype: string
- name: keywords
sequence: string
- name: ads_keywords
sequence: string
- name: read_count
dtype: int64
- name: doi
sequence: string
- name: authors
sequence: string
- name: aff
sequence: string
- name: cite_bibcodes
sequence: string
- name: ref_bibcodes
sequence: string
splits:
- name: train
num_bytes: 497970610
num_examples: 41195
download_size: 379674094
dataset_size: 497970610
configs:
- config_name: default
data_files:
- split: train
path: data/train-*