metadata
language:
- en
dataset_info:
features:
- name: text
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 1877736687
num_examples: 11615290
download_size: 1072595851
dataset_size: 1877736687
configs:
- config_name: default
data_files:
- split: train
path: data/train-*