metadata
dataset_info:
features:
- name: pubmed_id
dtype: string
- name: abstract
dtype: string
splits:
- name: train
num_bytes: 7119567
num_examples: 2963
download_size: 3777106
dataset_size: 7119567
configs:
- config_name: default
data_files:
- split: train
path: data/train-*