metadata
dataset_info:
features:
- name: text
dtype: string
- name: id
dtype: string
- name: token_count
dtype: int64
- name: corrupt
dtype: string
splits:
- name: train
num_bytes: 3812219177.6666665
num_examples: 385024
- name: test
num_bytes: 81111046.33333333
num_examples: 8192
download_size: 3223676991
dataset_size: 3893330224
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*