metadata
dataset_info:
features:
- name: query_id
dtype: int32
- name: answers
sequence: string
- name: passages
struct:
- name: is_selected
sequence: int32
- name: passage_text
sequence: string
- name: url
sequence: string
- name: query
dtype: string
- name: query_type
dtype: string
- name: wellFormedAnswers
sequence: 'null'
- name: ai_answers
dtype: string
- name: query_len
dtype: int64
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 85426944
num_examples: 20000
download_size: 41558640
dataset_size: 85426944
configs:
- config_name: default
data_files:
- split: train
path: data/train-*