metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 103158459
num_examples: 22939
download_size: 42323558
dataset_size: 103158459
configs:
- config_name: default
data_files:
- split: train
path: data/train-*