---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: labels
    sequence: int64
  splits:
  - name: train
    num_bytes: 8970605260
    num_examples: 336785
  - name: validation
    num_bytes: 89363780
    num_examples: 3355
  download_size: 2632990365
  dataset_size: 9059969040
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---