metadata
dataset_info:
features:
- name: index
dtype: int64
- name: q_id
dtype: string
- name: question
dtype: string
- name: best_answer
dtype: string
- name: all_answers
sequence: string
- name: num_answers
dtype: int64
- name: top_answers
sequence: string
- name: num_top_answers
dtype: int64
- name: context
dtype: string
- name: orig
dtype: string
- name: target
dtype: string
splits:
- name: train
num_bytes: 277046806
num_examples: 10000
- name: test
num_bytes: 34381247
num_examples: 1250
- name: validation
num_bytes: 34068900
num_examples: 1250
download_size: 214388763
dataset_size: 345496953
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*