metadata
dataset_info:
features:
- name: conversation
list:
- name: from
dtype: string
- name: value
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 38041275
num_examples: 47463
download_size: 20521038
dataset_size: 38041275
configs:
- config_name: default
data_files:
- split: train
path: data/train-*