metadata
dataset_info:
features:
- name: story
dtype: string
- name: generated_text_1
dtype: string
- name: generated_text_2
dtype: string
- name: mic_chosen
dtype: int64
- name: mar_chosen
dtype: int64
- name: ali_chosen
dtype: int64
splits:
- name: train
num_bytes: 84559
num_examples: 150
download_size: 56058
dataset_size: 84559
configs:
- config_name: default
data_files:
- split: train
path: data/train-*