metadata
dataset_info:
features:
- name: input
dtype: string
- name: actions
dtype: string
- name: num_blocks
dtype: int64
splits:
- name: train
num_bytes: 23398
num_examples: 100
download_size: 5433
dataset_size: 23398
configs:
- config_name: default
data_files:
- split: train
path: data/train-*