metadata
dataset_info:
features:
- name: context
dtype: string
- name: query
dtype: string
- name: target
dtype: string
splits:
- name: train
num_bytes: 243852549
num_examples: 1000000
- name: test
num_bytes: 1219361
num_examples: 5000
download_size: 196755731
dataset_size: 245071910
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*