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: 36288157
num_examples: 39069
- name: eval
num_bytes: 4052440
num_examples: 4341
download_size: 10638671
dataset_size: 40340597
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*