metadata
dataset_info:
features:
- name: prompt
dtype: large_string
- name: completion
dtype: large_string
splits:
- name: train
num_bytes: 87745059
num_examples: 23467
- name: variation
num_bytes: 34841
num_examples: 2000
download_size: 10248930
dataset_size: 87779900
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: variation
path: data/variation-*