metadata
dataset_info:
features:
- name: original_id
dtype: int64
- name: messages
list:
- name: role
dtype: string
- name: content
dtype: string
splits:
- name: train
num_bytes: 17493
num_examples: 3
download_size: 9813
dataset_size: 17493
configs:
- config_name: default
data_files:
- split: train
path: data/train-*