metadata
dataset_info:
features:
- name: input_ids
list: int32
- name: attention_mask
list: int8
- name: labels
list: int64
splits:
- name: train
num_bytes: 7173964660
num_examples: 1284369
download_size: 8831180453
dataset_size: 7173964660
configs:
- config_name: default
data_files:
- split: train
path: data/train-*