metadata
dataset_info:
features:
- name: message
dtype: string
- name: message_type
dtype: string
- name: message_id
dtype: int64
- name: conversation_id
dtype: int64
- name: cluster
dtype: float64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 9167321
num_examples: 21608
download_size: 4487797
dataset_size: 9167321
configs:
- config_name: default
data_files:
- split: train
path: data/train-*