metadata
dataset_info:
features:
- name: user_query
dtype: string
- name: ground_truth
sequence: string
- name: expected_steps
sequence: string
splits:
- name: train
num_bytes: 2728
num_examples: 20
download_size: 4241
dataset_size: 2728
configs:
- config_name: default
data_files:
- split: train
path: data/train-*