metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 105301517
num_examples: 4411
- name: test
num_bytes: 5562288
num_examples: 233
download_size: 110893427
dataset_size: 110863805
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*