metadata
dataset_info:
features:
- name: data_index_by_user
dtype: int64
- name: article
dtype: string
- name: answer
dtype: string
- name: question
dtype: string
- name: options
sequence: string
- name: input_ids
sequence:
sequence: int32
- name: token_type_ids
sequence:
sequence: int8
- name: attention_mask
sequence:
sequence: int8
- name: label
dtype: int64
splits:
- name: train
num_bytes: 170857353.97264022
num_examples: 15716
download_size: 55886476
dataset_size: 170857353.97264022
configs:
- config_name: default
data_files:
- split: train
path: data/train-*