metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: class
dtype: string
splits:
- name: train
num_bytes: 33930948
num_examples: 258037
- name: valid
num_bytes: 1434380
num_examples: 10752
download_size: 19778335
dataset_size: 35365328
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*