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: 38672620
num_examples: 39069
- name: eval
num_bytes: 4315670
num_examples: 4341
download_size: 11017876
dataset_size: 42988290
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*