metadata
license: apache-2.0
dataset_info:
features:
- name: answer
dtype: string
- name: reference
dtype: string
- name: extracted_pieces
sequence: string
splits:
- name: train
num_bytes: 647952505
num_examples: 5000
- name: validation
num_bytes: 128148693
num_examples: 1000
- name: test
num_bytes: 128148693
num_examples: 1000
download_size: 433880188
dataset_size: 904249891
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*