metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: rejected
dtype: string
- name: rejected_model_name
dtype: string
- name: accepted
dtype: string
- name: accepted_model_name
dtype: string
splits:
- name: train
num_bytes: 8850596
num_examples: 565
download_size: 3994041
dataset_size: 8850596
configs:
- config_name: default
data_files:
- split: train
path: data/train-*