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: 256138013.27959523
num_examples: 234258
- name: test
num_bytes: 28460143.720404785
num_examples: 26029
download_size: 87896869
dataset_size: 284598157
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*