Metadata-Version: 2.4
Name: bpe_knockout
Version: 2025.12.4
Summary: Implementation of BPE-knockout, a morphologically informed post-processing step for BPE tokenisers.
Author-email: Thomas Bauwens <thomas.bauwens@kuleuven.be>
Maintainer-email: Thomas Bauwens <thomas.bauwens@kuleuven.be>
License:                                  Apache License
                                   Version 2.0, January 2004
                                http://www.apache.org/licenses/
        
           TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
        
           1. Definitions.
        
              "License" shall mean the terms and conditions for use, reproduction,
              and distribution as defined by Sections 1 through 9 of this document.
        
              "Licensor" shall mean the copyright owner or entity authorized by
              the copyright owner that is granting the License.
        
              "Legal Entity" shall mean the union of the acting entity and all
              other entities that control, are controlled by, or are under common
              control with that entity. For the purposes of this definition,
              "control" means (i) the power, direct or indirect, to cause the
              direction or management of such entity, whether by contract or
              otherwise, or (ii) ownership of fifty percent (50%) or more of the
              outstanding shares, or (iii) beneficial ownership of such entity.
        
              "You" (or "Your") shall mean an individual or Legal Entity
              exercising permissions granted by this License.
        
              "Source" form shall mean the preferred form for making modifications,
              including but not limited to software source code, documentation
              source, and configuration files.
        
              "Object" form shall mean any form resulting from mechanical
              transformation or translation of a Source form, including but
              not limited to compiled object code, generated documentation,
              and conversions to other media types.
        
              "Work" shall mean the work of authorship, whether in Source or
              Object form, made available under the License, as indicated by a
              copyright notice that is included in or attached to the work
              (an example is provided in the Appendix below).
        
              "Derivative Works" shall mean any work, whether in Source or Object
              form, that is based on (or derived from) the Work and for which the
              editorial revisions, annotations, elaborations, or other modifications
              represent, as a whole, an original work of authorship. For the purposes
              of this License, Derivative Works shall not include works that remain
              separable from, or merely link (or bind by name) to the interfaces of,
              the Work and Derivative Works thereof.
        
              "Contribution" shall mean any work of authorship, including
              the original version of the Work and any modifications or additions
              to that Work or Derivative Works thereof, that is intentionally
              submitted to Licensor for inclusion in the Work by the copyright owner
              or by an individual or Legal Entity authorized to submit on behalf of
              the copyright owner. For the purposes of this definition, "submitted"
              means any form of electronic, verbal, or written communication sent
              to the Licensor or its representatives, including but not limited to
              communication on electronic mailing lists, source code control systems,
              and issue tracking systems that are managed by, or on behalf of, the
              Licensor for the purpose of discussing and improving the Work, but
              excluding communication that is conspicuously marked or otherwise
              designated in writing by the copyright owner as "Not a Contribution."
        
              "Contributor" shall mean Licensor and any individual or Legal Entity
              on behalf of whom a Contribution has been received by Licensor and
              subsequently incorporated within the Work.
        
           2. Grant of Copyright License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              copyright license to reproduce, prepare Derivative Works of,
              publicly display, publicly perform, sublicense, and distribute the
              Work and such Derivative Works in Source or Object form.
        
           3. Grant of Patent License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              (except as stated in this section) patent license to make, have made,
              use, offer to sell, sell, import, and otherwise transfer the Work,
              where such license applies only to those patent claims licensable
              by such Contributor that are necessarily infringed by their
              Contribution(s) alone or by combination of their Contribution(s)
              with the Work to which such Contribution(s) was submitted. If You
              institute patent litigation against any entity (including a
              cross-claim or counterclaim in a lawsuit) alleging that the Work
              or a Contribution incorporated within the Work constitutes direct
              or contributory patent infringement, then any patent licenses
              granted to You under this License for that Work shall terminate
              as of the date such litigation is filed.
        
           4. Redistribution. You may reproduce and distribute copies of the
              Work or Derivative Works thereof in any medium, with or without
              modifications, and in Source or Object form, provided that You
              meet the following conditions:
        
              (a) You must give any other recipients of the Work or
                  Derivative Works a copy of this License; and
        
              (b) You must cause any modified files to carry prominent notices
                  stating that You changed the files; and
        
              (c) You must retain, in the Source form of any Derivative Works
                  that You distribute, all copyright, patent, trademark, and
                  attribution notices from the Source form of the Work,
                  excluding those notices that do not pertain to any part of
                  the Derivative Works; and
        
              (d) If the Work includes a "NOTICE" text file as part of its
                  distribution, then any Derivative Works that You distribute must
                  include a readable copy of the attribution notices contained
                  within such NOTICE file, excluding those notices that do not
                  pertain to any part of the Derivative Works, in at least one
                  of the following places: within a NOTICE text file distributed
                  as part of the Derivative Works; within the Source form or
                  documentation, if provided along with the Derivative Works; or,
                  within a display generated by the Derivative Works, if and
                  wherever such third-party notices normally appear. The contents
                  of the NOTICE file are for informational purposes only and
                  do not modify the License. You may add Your own attribution
                  notices within Derivative Works that You distribute, alongside
                  or as an addendum to the NOTICE text from the Work, provided
                  that such additional attribution notices cannot be construed
                  as modifying the License.
        
              You may add Your own copyright statement to Your modifications and
              may provide additional or different license terms and conditions
              for use, reproduction, or distribution of Your modifications, or
              for any such Derivative Works as a whole, provided Your use,
              reproduction, and distribution of the Work otherwise complies with
              the conditions stated in this License.
        
           5. Submission of Contributions. Unless You explicitly state otherwise,
              any Contribution intentionally submitted for inclusion in the Work
              by You to the Licensor shall be under the terms and conditions of
              this License, without any additional terms or conditions.
              Notwithstanding the above, nothing herein shall supersede or modify
              the terms of any separate license agreement you may have executed
              with Licensor regarding such Contributions.
        
           6. Trademarks. This License does not grant permission to use the trade
              names, trademarks, service marks, or product names of the Licensor,
              except as required for reasonable and customary use in describing the
              origin of the Work and reproducing the content of the NOTICE file.
        
           7. Disclaimer of Warranty. Unless required by applicable law or
              agreed to in writing, Licensor provides the Work (and each
              Contributor provides its Contributions) on an "AS IS" BASIS,
              WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
              implied, including, without limitation, any warranties or conditions
              of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
              PARTICULAR PURPOSE. You are solely responsible for determining the
              appropriateness of using or redistributing the Work and assume any
              risks associated with Your exercise of permissions under this License.
        
           8. Limitation of Liability. In no event and under no legal theory,
              whether in tort (including negligence), contract, or otherwise,
              unless required by applicable law (such as deliberate and grossly
              negligent acts) or agreed to in writing, shall any Contributor be
              liable to You for damages, including any direct, indirect, special,
              incidental, or consequential damages of any character arising as a
              result of this License or out of the use or inability to use the
              Work (including but not limited to damages for loss of goodwill,
              work stoppage, computer failure or malfunction, or any and all
              other commercial damages or losses), even if such Contributor
              has been advised of the possibility of such damages.
        
           9. Accepting Warranty or Additional Liability. While redistributing
              the Work or Derivative Works thereof, You may choose to offer,
              and charge a fee for, acceptance of support, warranty, indemnity,
              or other liability obligations and/or rights consistent with this
              License. However, in accepting such obligations, You may act only
              on Your own behalf and on Your sole responsibility, not on behalf
              of any other Contributor, and only if You agree to indemnify,
              defend, and hold each Contributor harmless for any liability
              incurred by, or claims asserted against, such Contributor by reason
              of your accepting any such warranty or additional liability.
        
           END OF TERMS AND CONDITIONS
        
           APPENDIX: How to apply the Apache License to your work.
        
              To apply the Apache License to your work, attach the following
              boilerplate notice, with the fields enclosed by brackets "[]"
              replaced with your own identifying information. (Don't include
              the brackets!)  The text should be enclosed in the appropriate
              comment syntax for the file format. We also recommend that a
              file or class name and description of purpose be included on the
              same "printed page" as the copyright notice for easier
              identification within third-party archives.
        
           Copyright [yyyy] [name of copyright owner]
        
           Licensed under the Apache License, Version 2.0 (the "License");
           you may not use this file except in compliance with the License.
           You may obtain a copy of the License at
        
               http://www.apache.org/licenses/LICENSE-2.0
        
           Unless required by applicable law or agreed to in writing, software
           distributed under the License is distributed on an "AS IS" BASIS,
           WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
           See the License for the specific language governing permissions and
           limitations under the License.
        
Keywords: tokenizers,NLP,BPE,byte-pair encoding
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: OS Independent
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: transformers>=4.17.0
Requires-Dist: tokenizers>=0.13.2
Requires-Dist: datasets>=2.10.1
Requires-Dist: torch
Requires-Dist: requests
Requires-Dist: langcodes
Requires-Dist: language_data
Requires-Dist: numpy>=1.23.5
Requires-Dist: tqdm
Provides-Extra: github
Provides-Extra: tests
Requires-Dist: scipy; extra == "tests"
Dynamic: license-file

<img src="https://raw.githubusercontent.com/bauwenst/BPE-knockout/v2025.12.04/doc/logo.png">

Repo hosting all the code used for the NAACL 2024 paper [*"BPE-knockout: Pruning Pre-existing BPE Tokenisers with Backwards-compatible Morphological Semi-supervision"*](https://aclanthology.org/2024.naacl-long.324/),
including source code for BPE-knockout tokenisation.

Want to reproduce or extend the intrinsic evaluations of the paper? Read on!
Extrinsic evaluations are done with [RobBERT's](https://github.com/iPieter/RobBERT) framework. The pre-trained model
checkpoints are available on the [HuggingFace Hub](https://huggingface.co/collections/Bauwens/bpe-knockout-660be8a33336a7e1289be624).

Table of contents:
- [Installation](#installing)
- Usage:
  - [HuggingFace compatibility](#huggingface-compatibility)
  - [Saving and loading](#more-usage-examples)
- Paper results:
  - [Running experiments from the paper](#running-experiments-from-the-paper)
  - [Using your own data](#using-your-own-data)
- [Data licenses](#data-licenses)
- [BibTeX citation](#citation)

## HuggingFace compatibility
Are you used to working with the HuggingFace suite for language modelling and tokenisation? No problem!
You can incorporate BPE-knockout anywhere you're already using a BPE tokeniser loaded from HuggingFace, 
with only 2 extra imports and 2 more lines of code. For example, if you're using `roberta-base`'s English tokeniser, 
you would run:
```python
# Load HuggingFace object
from transformers import AutoTokenizer
hf_bpe_tokeniser = AutoTokenizer.from_pretrained("roberta-base")

# Construct TkTkT object
from tktkt.models.bpe.knockout import BPEKnockout
tktkt_bpek_tokeniser = BPEKnockout.fromHuggingFace(hf_bpe_tokeniser, "English")

# Convert back to HuggingFace
from tktkt.interfaces.huggingface import TktktToHuggingFace
hf_bpek_tokeniser = TktktToHuggingFace(tktkt_bpek_tokeniser, specials_from=hf_bpe_tokeniser)
```
The resulting object is indeed a HuggingFace tokeniser, but internally it works using BPE-knockout.

## Installing
### Minimal package
If you are only interested in using the BPE-knockout package (including our English, German and Dutch BPE tokenisers and
the respective morphological data loaders, but **not** including corpus word counts) and not in running the experiments
from the paper, you likely just want to run:
```shell
pip install "bpe_knockout[github] @ git+https://github.com/bauwenst/BPE-knockout"
```
As shown in the above example, user-friendly encapsulations for BPE-knockout are provided by the [TkTkT package](https://github.com/bauwenst/TkTkT),
which may be more interesting to you than the core algorithm and configuration code which is provided here. In any case, installing
either package will install the other automatically anyway.

### Full experiments, editable code
If you want to run experiments from the paper and/or have access to the word count files, this means you want to download
everything in this repository and tell Python to use the folder into which you cloned for the package code, rather than
copying the code to your global or virtual `site-packages` directory. In that case, run:
```shell
git clone https://github.com/bauwenst/BPE-knockout
cd BPE-knockout
pip install -e .[github]
```
*Warning*:
- If you're using conda or venv, don't forget to activate your environment before running any calls to `pip install`.
- If you have an editable installation of my other packages `TkTkT` and/or `Fiject` and would like to keep it, do *not* include the `[github]` suffix.

## More usage examples
### Saving and loading tokeniser after knockout
Although knockout takes under 3 minutes to complete and lasts as long as the tokeniser's lifetime, you may still want to
cache the resulting BPE-knockout tokeniser for using it again without re-doing knockout. Starting from a HuggingFace tokeniser,
that would look like this:
```python
from pathlib import Path
from transformers import AutoTokenizer
from tktkt.models.bpe.knockout import BPEKnockout
from tktkt.models.huggingface.wrapper import HuggingFacePreprocessor

# Load old tokeniser and apply knockout.
hf_base = AutoTokenizer.from_pretrained("roberta-base")
tktkt_knockout = BPEKnockout.fromHuggingFace(hf_base, language="English")

# Save new tokeniser and reload it.
save_path = tktkt_knockout.save(folder=Path(".") / "roberta-knockout")
tktkt_knockout_loaded = BPEKnockout.load(save_path, preprocessor=HuggingFacePreprocessor(hf_base))
```
The objects `tktkt_knockout` and `tktkt_knockout_loaded` have the exact same internals. (If you want to give them the
HuggingFace interface, wrap them with a call to `TktktToHuggingFace`.)

Notice how `.load()` requires a preprocessor. That's because `.save()` does not store the preprocessor with the tokeniser,
as it assumes that you already have a code snippet that loads the preprocessor. In this case, because the original tokeniser
was a HuggingFace tokeniser, its preprocessor was a HuggingFace preprocessor as well.

### Loading from the HuggingFace hub
The BPE-knockout tokeniser used for the (continued) pre-training of language models in the paper were saved as above and
uploaded to the HuggingFace hub. Rather than downloading them manually, you can load them with a `from_pretrained` call.
For convenience, there is one that has the TkTkT interface and one that converts it to the HuggingFace interface:
```python
from tktkt.models.bpe.knockout import BPEKnockout

dutch_bpe_knockout_hf    = BPEKnockout.from_pretrained("Bauwens/RoBERTa-nl_BPE_30k_BPE-knockout_9k")
dutch_bpe_knockout_tktkt = BPEKnockout.from_pretrained_tktkt("Bauwens/RoBERTa-nl_BPE_30k_BPE-knockout_9k")
```

## Running experiments from the paper
Given that you have an editable install, follow these steps to reproduce the paper results:
1. Unzip the `.rar` file under `data/compressed/`.
2. Run `py tst/main.py` or `python tst/main.py` in a terminal.

## Using your own data
It is possible to use other datasets (even other languages) than the ones used for the paper.
All files read by the package are declared in a globally accessible `ProjectConfig` object. We ship three default
`ProjectConfig`s with the package for English, German and Dutch (see the `bpe_knockout.project.config` file).

To add a config, you'll need the following files:
   1. A dataset of morphological decompositions. BPE-knockout is built on top of [MoDeST](https://github.com/bauwenst/MoDeST)
      for supplying morphological data, so your data format must be compatible with those supported by MoDeST. (Otherwise,
      you can always write your own class.)
   2. *Optional:* if you want to run all experiments, you also need a tab-separated file of words and their frequencies from a sufficiently large corpus (morphologies without such a frequency will get frequency 1, and words that don't have a morphology will be ignored);
   3. *Optional:* if you don't want to generate a new BPE tokeniser from your word counts, the file(s) that specify your 
      existing BPE tokeniser.

Now write a function akin to `setupEnglish()` to return your new config. If you want to run the experiments on these new data, 
import it in `tst/main.py`. If you want to just use these new data for knockout, import your function in `tktkt.models.bpe.knockout`
and add it to the `CONFIGS` dictionary.

## Data licenses
All data is included in the repo, because it is obtainable for free elsewhere and free of license too.
- Morphological decompositions were derived from [WebCelex at the Max Plank Institute](http://celex.mpi.nl/).
- Language modelling data is derived from [OSCAR on HuggingFace](https://huggingface.co/datasets/oscar).

## Citation
If you use BPE-knockout in your own work, cite the paper using e.g.:
```bibtex
@inproceedings{bauwens-delobelle-2024-bpe,
    title = "{BPE}-knockout: Pruning Pre-existing {BPE} Tokenisers with Backwards-compatible Morphological Semi-supervision",
    author = "Bauwens, Thomas  and  Delobelle, Pieter",
    booktitle = "Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
    month = jun,
    year = "2024",
    address = "Mexico City, Mexico",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.naacl-long.324",
    pages = "5810--5832"
}
```
