metadata
dataset_info:
features:
- name: id
dtype: string
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answer_text
dtype: string
- name: answer_start
dtype: float64
- name: language
dtype: string
- name: negative_samples
sequence: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 659667586
num_examples: 147045
download_size: 60361682
dataset_size: 659667586
configs:
- config_name: default
data_files:
- split: train
path: data/train-*