metadata
dataset_info:
features:
- name: question
dtype: string
- name: language
dtype: string
- name: golds
struct:
- name: answer_start
sequence: int64
- name: answer_text
sequence: string
- name: context
dtype: string
- name: seq_id
dtype: string
splits:
- name: train
num_bytes: 21022889
num_examples: 29868
- name: validation
num_bytes: 2616173
num_examples: 3712
download_size: 16292808
dataset_size: 23639062