metadata
dataset_info:
features:
- name: video
sequence:
sequence:
sequence:
sequence: float32
splits:
- name: train
num_bytes: 4993216000
num_examples: 1000
- name: val
num_bytes: 499321600
num_examples: 100
download_size: 154487073
dataset_size: 5492537600
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*