metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
- name: id
dtype: string
- name: source
dtype: string
- name: category
dtype: string
- name: subcategory
dtype: string
splits:
- name: train
num_bytes: 2890039
num_examples: 4000
- name: test
num_bytes: 148130
num_examples: 200
download_size: 1155602
dataset_size: 3038169
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*