---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: labels
    sequence: int64
  splits:
  - name: train
    num_bytes: 17609664600
    num_examples: 1321650
  - name: validation
    num_bytes: 914479416
    num_examples: 68634
  download_size: 5516198279
  dataset_size: 18524144016
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---