metadata
dataset_info:
features:
- name: text
dtype: string
- name: image
struct:
- name: bytes
dtype: binary
- name: endings
list: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 81341892
num_examples: 1024
download_size: 22346162
dataset_size: 81341892
configs:
- config_name: default
data_files:
- split: train
path: data/train-*