metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: embedding
sequence: float64
- name: result
struct:
- name: ds_index
dtype: int64
- name: least_similar_dg_indices
sequence: int64
splits:
- name: train
num_bytes: 52986082
num_examples: 7473
download_size: 41248185
dataset_size: 52986082
configs:
- config_name: default
data_files:
- split: train
path: data/train-*