---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  splits:
  - name: train
    num_bytes: 928847774.5358719
    num_examples: 7635694
  - name: test
    num_bytes: 1216454.9476915547
    num_examples: 10000
  download_size: 523969256
  dataset_size: 930064229.4835634
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---

256トークンで切った。