metadata
dataset_info:
features:
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 89534315
num_examples: 251998
- name: test
num_bytes: 11192313
num_examples: 31500
- name: validation
num_bytes: 11182215
num_examples: 31500
download_size: 64896112
dataset_size: 111908843
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*