metadata
dataset_info:
features:
- name: title
dtype: string
- name: url
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: answer_start
dtype: 'null'
- name: id
dtype: string
- name: context_tok_len
dtype: int64
- name: question_list
sequence: string
- name: document
dtype: string
- name: summary
dtype: string
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 334009605
num_examples: 9583
download_size: 151945756
dataset_size: 334009605
configs:
- config_name: default
data_files:
- split: train
path: data/train-*