metadata
dataset_info:
features:
- name: qid
dtype: string
- name: question
dtype: string
- name: answer_candidates
sequence: string
- name: correct_answer_index
dtype: int64
- name: input_ids
sequence:
sequence: int32
- name: token_type_ids
sequence:
sequence: int8
- name: attention_mask
sequence:
sequence: int8
splits:
- name: train
num_bytes: 80377569
num_examples: 13061
- name: validation
num_bytes: 6438977
num_examples: 995
- name: test
num_bytes: 6443893
num_examples: 997
download_size: 16067303
dataset_size: 93260439
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*