metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: source
dtype: string
- name: instruction
dtype: string
- name: response
dtype: string
splits:
- name: train
num_bytes: 68530
num_examples: 15
download_size: 47705
dataset_size: 68530
configs:
- config_name: default
data_files:
- split: train
path: data/train-*