metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 30966.02564102564
num_examples: 35
- name: test
num_bytes: 3538.974358974359
num_examples: 4
download_size: 26718
dataset_size: 34505
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*