---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: chunk_starts
    sequence: int64
  - name: length
    dtype: int64
  splits:
  - name: train
    num_bytes: 269349008
    num_examples: 1024
  - name: validation
    num_bytes: 8416488
    num_examples: 32
  - name: test
    num_bytes: 33667920
    num_examples: 128
  download_size: 160304458
  dataset_size: 311433416
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---