metadata
dataset_info:
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: text
dtype: string
- name: filename
dtype: string
- name: _image
dtype: string
splits:
- name: train
num_bytes: 42465505
num_examples: 4000
download_size: 40681500
dataset_size: 42465505
configs:
- config_name: default
data_files:
- split: train
path: data/train-*