metadata
dataset_info:
features:
- name: image
dtype: image
- name: filename
dtype: string
- name: words
dtype: string
splits:
- name: train
num_bytes: 3683256
num_examples: 101
download_size: 3664632
dataset_size: 3683256
configs:
- config_name: default
data_files:
- split: train
path: data/train-*