metadata
dataset_info:
features:
- name: image
dtype: image
- name: 'Unnamed: 0'
dtype: int64
- name: text
dtype: string
splits:
- name: train
num_bytes: 333595637.208
num_examples: 20638
download_size: 343286724
dataset_size: 333595637.208
configs:
- config_name: default
data_files:
- split: train
path: data/train-*