metadata
dataset_info:
features:
- name: id
dtype: string
- name: conversations
list:
- name: role
dtype: string
- name: value
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 46327520
num_examples: 20000
download_size: 24763407
dataset_size: 46327520
configs:
- config_name: default
data_files:
- split: train
path: data/train-*