metadata
dataset_info:
features:
- name: instruction_seed
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 37080197
num_examples: 46000
download_size: 19671981
dataset_size: 37080197
configs:
- config_name: default
data_files:
- split: train
path: data/train-*