metadata
dataset_info:
features:
- name: source_id
dtype: int64
- name: question
dtype: string
- name: response
dtype: string
- name: metadata
struct:
- name: source
sequence: string
splits:
- name: train
num_bytes: 125049216
num_examples: 46860
download_size: 64816414
dataset_size: 125049216
configs:
- config_name: default
data_files:
- split: train
path: data/train-*