metadata
dataset_info:
features:
- name: image
dtype: binary
- name: label
dtype: int64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 29228528944
num_examples: 81167
download_size: 29229537168
dataset_size: 29228528944
configs:
- config_name: default
data_files:
- split: train
path: data/train-*