metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: expert_name
dtype: string
- name: task_eval_on
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 19490
num_examples: 440
download_size: 12510
dataset_size: 19490