metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: label
dtype: int64
- name: split
dtype: string
splits:
- name: train
num_bytes: 2206914500.36
num_examples: 9988
- name: test
num_bytes: 600218593.152
num_examples: 2672
download_size: 2692003047
dataset_size: 2807133093.512
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*