---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  splits:
  - name: train
    num_bytes: 26575097543
    num_examples: 6000000
  download_size: 8682781157
  dataset_size: 26575097543
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---