metadata
dataset_info:
features:
- name: index
dtype: string
- name: text
dtype: string
- name: doi
dtype: string
splits:
- name: train
num_bytes: 6874262385.655861
num_examples: 5000000
- name: validation
num_bytes: 34354751.90185088
num_examples: 25000
download_size: 4045472129
dataset_size: 6908617137.557712
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*