metadata
dataset_info:
features:
- name: conversation
list:
- name: from
dtype: string
- name: value
dtype: string
- name: sys_message
dtype: string
- name: tkn_len
dtype: int64
splits:
- name: train
num_bytes: 1145395609.5420778
num_examples: 396955
download_size: 559019729
dataset_size: 1145395609.5420778
configs:
- config_name: default
data_files:
- split: train
path: data/train-*