metadata
dataset_info:
features:
- name: paper_id
dtype: string
- name: items
list: string
- name: gt
dtype: float64
splits:
- name: train
num_bytes: 16667731
num_examples: 2305
- name: validation
num_bytes: 4797950
num_examples: 668
download_size: 11307007
dataset_size: 21465681
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*