metadata
dataset_info:
features:
- name: messages
list:
- name: role
dtype: string
- name: content
dtype: string
splits:
- name: train
num_bytes: 15105137
num_examples: 3970
- name: validation
num_bytes: 780532
num_examples: 210
download_size: 7892992
dataset_size: 15885669
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*