metadata
dataset_info:
features:
- name: questions
dtype: string
- name: answers
dtype: string
- name: retrieved
sequence: string
- name: new_chunks
sequence: string
- name: chunker_ids
sequence: string
splits:
- name: dev
num_bytes: 2978956171
num_examples: 8757
download_size: 1494745908
dataset_size: 2978956171
configs:
- config_name: default
data_files:
- split: dev
path: data/dev-*