metadata
dataset_info:
features:
- name: query
dtype: string
- name: answer
dtype: string
- name: question_type
dtype: string
- name: evidence_list
list:
- name: author
dtype: string
- name: category
dtype: string
- name: fact
dtype: string
- name: published_at
dtype: string
- name: source
dtype: string
- name: title
dtype: string
- name: url
dtype: string
- name: embedding_ada
sequence: float64
- name: embedding_small
sequence: float64
- name: embedding_large
sequence: float64
splits:
- name: train
num_bytes: 97838126
num_examples: 2556
download_size: 72326469
dataset_size: 97838126
configs:
- config_name: default
data_files:
- split: train
path: data/train-*