metadata
dataset_info:
features:
- name: global_id
dtype: string
- name: messages
list:
- name: role
dtype: string
- name: content
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 203991994
num_examples: 9538
download_size: 203853921
dataset_size: 203991994
configs:
- config_name: default
data_files:
- split: train
path: data/train-*