metadata
dataset_info:
features:
- name: compressed_context
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 77686537
num_examples: 9292
download_size: 20365531
dataset_size: 77686537
configs:
- config_name: default
data_files:
- split: train
path: data/train-*