metadata
dataset_info:
features:
- name: instruction_seed
dtype: string
- name: response_seed
dtype: string
splits:
- name: train
num_bytes: 20367008
num_examples: 90800
download_size: 701490
dataset_size: 20367008
configs:
- config_name: default
data_files:
- split: train
path: data/train-*