metadata
dataset_info:
features:
- name: cache_key
dtype: string
- name: dataset_name
dtype: string
- name: frame_skip
dtype: int64
- name: num_observation_images
dtype: int64
- name: valid_indices
sequence: int64
- name: total_frames
dtype: int64
- name: created_at
dtype: string
splits:
- name: train
num_bytes: 46768629
num_examples: 19
download_size: 23681609
dataset_size: 46768629
configs:
- config_name: default
data_files:
- split: train
path: data/train-*