metadata
dataset_info:
features:
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: question
dtype: string
splits:
- name: train
num_bytes: 204766191
num_examples: 160800
download_size: 120808924
dataset_size: 204766191
configs:
- config_name: default
data_files:
- split: train
path: data/train-*