metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: src
dtype: string
splits:
- name: train
num_bytes: 1231378567
num_examples: 1228285
download_size: 669677299
dataset_size: 1231378567
configs:
- config_name: default
data_files:
- split: train
path: data/train-*