metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: idx
dtype: int64
- name: audio
sequence: int16
- name: video
sequence:
sequence:
sequence: uint8
- name: label
dtype: string
splits:
- name: train
num_bytes: 824374107
num_examples: 1321
download_size: 677311360
dataset_size: 824374107