metadata
dataset_info:
features:
- name: language
dtype: string
- name: topic
dtype: string
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 172791772
num_examples: 221000
download_size: 24922703
dataset_size: 172791772
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: apache-2.0