metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: completion
dtype: string
- name: text
dtype: string
- name: char_length
dtype: int64
splits:
- name: train
num_bytes: 54645165
num_examples: 900000
- name: validation
num_bytes: 6071685
num_examples: 100000
download_size: 29827729
dataset_size: 60716850
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*