metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 320337415956
num_examples: 9827644
- name: validation
num_bytes: 17665801264
num_examples: 541974
- name: test
num_bytes: 17764289532
num_examples: 544975
download_size: 63566585653
dataset_size: 355767506752
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*