metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: seed_prompt
dtype: string
- name: generation
dtype: string
splits:
- name: train
num_bytes: 16359316
num_examples: 2992
download_size: 4758599
dataset_size: 16359316
configs:
- config_name: default
data_files:
- split: train
path: data/train-*