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: 38772282.58722768
num_examples: 50970
download_size: 20538245
dataset_size: 38772282.58722768