metadata
dataset_info:
features:
- name: instruction_seed
dtype: string
- name: found_ngrams
sequence:
sequence: int64
splits:
- name: train
num_bytes: 33343828.622209772
num_examples: 15027
download_size: 20044782
dataset_size: 33343828.622209772
configs:
- config_name: default
data_files:
- split: train
path: data/train-*