metadata
dataset_info:
features:
- name: prompt
list:
- name: content
dtype: string
- name: role
dtype: string
- name: response
dtype: string
- name: dataset
dtype: string
splits:
- name: train
num_bytes: 4177418
num_examples: 1607
download_size: 898703
dataset_size: 4177418
configs:
- config_name: default
data_files:
- split: train
path: data/train-*