metadata
dataset_info:
features:
- name: user
dtype: string
- name: response
dtype: string
splits:
- name: train
num_bytes: 200
num_examples: 20
download_size: 1164
dataset_size: 200
configs:
- config_name: default
data_files:
- split: train
path: data/train-*