metadata
dataset_info:
features:
- name: schema
dtype: string
- name: question
dtype: string
- name: rejected
dtype: string
- name: chosen
dtype: string
- name: weight
dtype: float64
splits:
- name: train
num_bytes: 250333739.2658651
num_examples: 234216
- name: test
num_bytes: 27815928.734134898
num_examples: 26025
download_size: 87415316
dataset_size: 278149668
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*