metadata
dataset_info:
features:
- name: episode_index
dtype: int64
- name: task_embeddings
sequence: float64
splits:
- name: train
num_bytes: 5180
num_examples: 5
download_size: 6149
dataset_size: 5180
configs:
- config_name: default
data_files:
- split: train
path: data/train-*