metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 4182722
num_examples: 1896
- name: validation
num_bytes: 998040
num_examples: 450
- name: test
num_bytes: 6156650
num_examples: 2766
download_size: 1581048
dataset_size: 11337412
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*