metadata
dataset_info:
features:
- name: conversation
list:
- name: content
dtype: string
- name: do_train
dtype: bool
- name: role
dtype: string
splits:
- name: train
num_bytes: 9428391
num_examples: 4536
download_size: 3208011
dataset_size: 9428391
