metadata
dataset_info:
features:
- name: ngram
sequence: string
splits:
- name: train
num_bytes: 120415667683
num_examples: 10404628
download_size: 25258393258
dataset_size: 120415667683
configs:
- config_name: default
data_files:
- split: train
path: data/train-*