metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: target
dtype: string
- name: meta_path
list: int64
- name: meta_rules
list: int64
- name: index
dtype: int64
splits:
- name: train
num_bytes: 75700000
num_examples: 100000
download_size: 15115022
dataset_size: 75700000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*