metadata
dataset_info:
features:
- name: label
sequence: int8
- name: label_as_text
dtype: string
- name: vae_output
sequence: int8
splits:
- name: train
num_bytes: 1322709708
num_examples: 320292
download_size: 1312907245
dataset_size: 1322709708
configs:
- config_name: default
data_files:
- split: train
path: data/train-*