metadata
dataset_info:
features:
- name: Numeric_ID
dtype: int64
- name: Prompt
dtype: string
- name: Answer_A
dtype: string
- name: Answer_B
dtype: string
- name: Category
dtype: string
- name: answers
sequence: string
- name: chosen
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train_prefs
num_bytes: 3678761.0847844705
num_examples: 596
download_size: 2334787
dataset_size: 3678761.0847844705
configs:
- config_name: default
data_files:
- split: train_prefs
path: data/train_prefs-*