---
dataset_info:
  features:
  - name: language
    dtype: int64
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: depths
    sequence: int64
  - name: sibling_indices
    sequence: int64
  - name: tree_attention_mask
    sequence: int64
  splits:
  - name: train
    num_bytes: 2380160000.0
    num_examples: 160000
  - name: test
    num_bytes: 169512020.0
    num_examples: 11395
  - name: validation
    num_bytes: 164305420.0
    num_examples: 11045
  download_size: 40720475
  dataset_size: 2713977440.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: validation
    path: data/validation-*
---