metadata
dataset_info:
features:
- name: id
dtype: string
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answers
sequence:
- name: text
dtype: string
- name: answer_start
dtype: int32
- name: text
dtype: string
- name: answer_embedding
sequence: float64
splits:
- name: train
num_bytes: 122527254
num_examples: 8760
- name: test
num_bytes: 14707545
num_examples: 1057
download_size: 92115032
dataset_size: 137234799
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*