metadata
dataset_info:
features:
- name: user_input
dtype: string
- name: retrieved_contexts
sequence: string
- name: reference
dtype: string
splits:
- name: train
num_bytes: 37047
num_examples: 25
download_size: 14000
dataset_size: 37047
configs:
- config_name: default
data_files:
- split: train
path: data/train-*