metadata
dataset_info:
features:
- name: rejected_score
dtype: 'null'
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: chosen_score
dtype: 'null'
splits:
- name: train
num_bytes: 195208028
num_examples: 92316
- name: test
num_bytes: 48595565
num_examples: 23080
download_size: 137459167
dataset_size: 243803593
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*