metadata
dataset_info:
features:
- name: system
dtype: string
- name: question
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: hash
dtype: string
splits:
- name: train
num_bytes: 35797790
num_examples: 11537
download_size: 20824340
dataset_size: 35797790
configs:
- config_name: default
data_files:
- split: train
path: data/train-*