metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: prompt
dtype: string
- name: completion
dtype: string
- name: source
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 3294652388.329473
num_examples: 18964
download_size: 476508613
dataset_size: 3294652388.329473