metadata
dataset_info:
features:
- name: image_id
dtype: string
- name: subject_id
dtype: string
- name: study_id
dtype: string
- name: caption
dtype: string
splits:
- name: train
num_bytes: 20075822
num_examples: 50000
download_size: 6202101
dataset_size: 20075822
configs:
- config_name: default
data_files:
- split: train
path: data/train-*