metadata
dataset_info:
features:
- name: correct
dtype: string
- name: variant
dtype: string
- name: type
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 2460602
num_examples: 35598
download_size: 424415
dataset_size: 2460602
configs:
- config_name: default
data_files:
- split: train
path: data/train-*