metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: prompt_id
dtype: string
- name: category
dtype: string
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 16510442
num_examples: 9500
- name: validation
num_bytes: 888237
num_examples: 500
download_size: 11045341
dataset_size: 17398679
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*