metadata
dataset_info:
features:
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
- name: model_name
dtype: string
splits:
- name: train
num_bytes: 14789631
num_examples: 4380
download_size: 4480944
dataset_size: 14789631
configs:
- config_name: default
data_files:
- split: train
path: data/train-*