metadata
dataset_info:
features:
- name: inst
dtype: string
- name: good
dtype: string
- name: bad
dtype: string
splits:
- name: train
num_bytes: 97733120
num_examples: 34520
download_size: 16762622
dataset_size: 97733120
configs:
- config_name: default
data_files:
- split: train
path: data/train-*