metadata
dataset_info:
features:
- name: question
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 15653431
num_examples: 19534
- name: eval
num_bytes: 1733433
num_examples: 2171
download_size: 4798903
dataset_size: 17386864
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*