metadata
dataset_info:
features:
- name: image
dtype: image
- name: caption
dtype: string
- name: caption_part
dtype: string
- name: caption_partiel
dtype: string
splits:
- name: train
num_bytes: 19582741
num_examples: 154
download_size: 19332063
dataset_size: 19582741
configs:
- config_name: default
data_files:
- split: train
path: data/train-*