metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 1744822228
num_examples: 261671
- name: test
num_bytes: 160172028
num_examples: 24021
download_size: 284188358
dataset_size: 1904994256
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*