metadata
dataset_info:
features:
- name: prompts
dtype: string
- name: responses_1
dtype: string
- name: responses_2
dtype: string
- name: responses_merged
dtype: string
splits:
- name: train
num_bytes: 1631670
num_examples: 4000
download_size: 499953
dataset_size: 1631670
configs:
- config_name: default
data_files:
- split: train
path: data/train-*