metadata
dataset_info:
features:
- name: question
dtype: string
- name: language
dtype: string
- name: context
dtype: string
- name: seq_id
dtype: string
- name: golds
struct:
- name: answer_start
sequence: int64
- name: answer_text
sequence: string
splits:
- name: train
num_bytes: 32809511
num_examples: 129290
- name: validation
num_bytes: 4034498
num_examples: 15801
download_size: 17092210
dataset_size: 36844009