metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: unit
dtype: string
- name: book title
dtype: string
- name: audience
dtype: string
- name: completion
dtype: string
- name: token_length
dtype: int64
splits:
- name: train
num_bytes: 0
num_examples: 0
download_size: 1602
dataset_size: 0
configs:
- config_name: default
data_files:
- split: train
path: data/train-*