metadata
dataset_info:
features:
- name: text
dtype: string
- name: id
dtype: string
- name: url
dtype: string
- name: caption
dtype: string
- name: width
dtype: int64
- name: height
dtype: int64
- name: mime_type
dtype: string
- name: hash
dtype: string
- name: license
dtype: string
- name: source
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 115802308695.419
num_examples: 69999
download_size: 171546251609
dataset_size: 115802308695.419
configs:
- config_name: default
data_files:
- split: train
path: data/train-*