metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: chosen
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected
list:
- name: content
dtype: string
- name: role
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 603744333.9991473
num_examples: 184597
- name: test
num_bytes: 6099683.000852721
num_examples: 1865
download_size: 344241874
dataset_size: 609844017
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*