metadata
dataset_info:
features:
- name: conversation
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 7096640439
num_examples: 4110641
download_size: 3467280679
dataset_size: 7096640439
configs:
- config_name: default
data_files:
- split: train
path: data/train-*