metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 34853038872
num_examples: 327234
- name: validation
num_bytes: 3798075280
num_examples: 35660
- name: test
num_bytes: 4000546988
num_examples: 37561
download_size: 5539508167
dataset_size: 42651661140
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*