---
dataset_info:
  features:
  - name: text_id
    dtype: int64
  - name: chunk_id
    dtype: int64
  - name: pixel_values
    sequence:
      sequence: int64
  - name: num_patches
    dtype: int64
  - name: text
    dtype: string
  - name: tok_grapheme
    sequence: int64
  - name: tok_llama2
    sequence: int64
  - name: tok_komodo
    sequence: int64
  - name: tok_mt5
    sequence: int64
  splits:
  - name: train
    num_bytes: 17824895625
    num_examples: 293933
  - name: test
    num_bytes: 13297452
    num_examples: 823
  download_size: 517297380
  dataset_size: 17838193077
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---