metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: selected_model
dtype: string
- name: pair
dtype: string
- name: submission_time
dtype: string
splits:
- name: train
num_bytes: 3276
num_examples: 20
download_size: 2655
dataset_size: 3276
configs:
- config_name: default
data_files:
- split: train
path: data/train-*