metadata
dataset_info:
features:
- name: features
sequence: float32
- name: labels
dtype: string
splits:
- name: train
num_bytes: 4142196436
num_examples: 16560
download_size: 4000299254
dataset_size: 4142196436