metadata
dataset_info:
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 668239316.1
num_examples: 23436
- name: test
num_bytes: 74248812.9
num_examples: 2604
download_size: 190873523
dataset_size: 742488129
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*