---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: labels
    sequence: int64
  splits:
  - name: train
    num_bytes: 16990579172
    num_examples: 319012
  - name: test
    num_bytes: 4400128186
    num_examples: 82616
  download_size: 503505301
  dataset_size: 21390707358
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---