metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype: int64
- name: pixel_values
sequence:
sequence:
sequence: float32
splits:
- name: train
num_bytes: 29296386
num_examples: 48
download_size: 8261484
dataset_size: 29296386
configs:
- config_name: default
data_files:
- split: train
path: data/train-*