metadata
dataset_info:
features:
- name: generated
list:
- name: content
dtype: string
- name: role
dtype: string
- name: real
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 82544374.21505675
num_examples: 19426
- name: test
num_bytes: 9173957.784943247
num_examples: 2159
download_size: 51274646
dataset_size: 91718332
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*