metadata
dataset_info:
features:
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 4118938
num_examples: 772
- name: test
num_bytes: 518673
num_examples: 100
download_size: 2038755
dataset_size: 4637611
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*