metadata
dataset_info:
features:
- name: image
dtype: image
- name: pixel_values
sequence:
sequence:
sequence: float32
splits:
- name: train
num_bytes: 1764742614
num_examples: 894
download_size: 350336236
dataset_size: 1764742614
configs:
- config_name: default
data_files:
- split: train
path: data/train-*