metadata
dataset_info:
features:
- name: text
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 7206574781
num_examples: 203037
download_size: 3381871729
dataset_size: 7206574781
configs:
- config_name: default
data_files:
- split: train
path: data/train-*