metadata
dataset_info:
features:
- name: instruction_seed
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 774635827
num_examples: 3005927
download_size: 402206463
dataset_size: 774635827
configs:
- config_name: default
data_files:
- split: train
path: data/train-*