metadata
dataset_info:
features:
- name: text
dtype: string
- name: len_tokens
dtype: int64
splits:
- name: train
num_bytes: 44143916
num_examples: 48016
download_size: 8526144
dataset_size: 44143916
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
hist len_tokens
describe.
percentil.
