metadata
dataset_info:
features:
- name: id
dtype: int64
- name: messages
list:
- name: role
dtype: string
- name: content
dtype: string
splits:
- name: train
num_bytes: 14288072
num_examples: 8402
download_size: 7596816
dataset_size: 14288072
configs:
- config_name: default
data_files:
- split: train
path: data/train-*