RUKOPYS Leaderboard

RUKOPYS Leaderboard

Ranks OCR systems on reading handwritten Ukrainian, scored on whole pages from the RUKOPYS test split. Researchers and model builders can compare runs right in the browser.

Made by UkrainianCatholicUniversity on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This is a leaderboard page for machine reading of handwritten Ukrainian. It ranks OCR systems on whole document pages from the public test split of the RUKOPYS dataset. The numbers come from a separate results dataset, so a run merged there appears on the page with no change to the Space.

Who it suits: It suits researchers and engineers who build or compare text recognition for handwriting, and anyone who wants to see how systems do on Ukrainian pages.

Industry | Who uses it | How they use it | Example
Machine learning research | OCR researchers | Compare their model with other runs on the same test split | A lab checks where a new handwriting model lands among earlier runs
Digital archives | Digitization teams | Read the results before choosing a system for scanned handwritten pages | An archive team reviews the ranking before it picks an OCR system
Software | Developers | Pick an OCR system for a product that reads Ukrainian handwriting | A developer looks at the ranking before testing a model
Education | Students and teachers | Use it as a worked example of a public benchmark | A class studies how a leaderboard presents results

You can skip it if you work only with printed text or with languages other than Ukrainian.

Where to use it: You use it in these places.

• A web browser, on the Space's page on Hugging Face
• Windows, macOS or Linux, after you copy the files and open index.html in a browser

Use case | Who | What they get back
Compare results | Researchers | The results of OCR runs on the test split
See a new run | Maintainers | A merged run shows up on the page with no change to the Space
Review a benchmark | Students and developers | A view of how handwriting OCR is scored on whole pages

How to set it up: The online page needs nothing except a web browser. A local copy needs Python and pip.

• A web browser
• Python with pip, only for the local copy

1. Open the Space's page on Hugging Face to use it online, with nothing to install.
2. For a local copy, install the Hugging Face Hub package.
3. Download the Space's files into a folder.
4. Move into that folder and open index.html in a web browser.

pip install -U huggingface_hub
hf download UkrainianCatholicUniversity/rukopys-leaderboard --repo-type space --local-dir rukopys-leaderboard
cd rukopys-leaderboard

How to use it: Open the page and read the results. It loads its data from the results dataset, so you see the runs merged there. With a local copy, open index.html in your browser after the download.

hf download UkrainianCatholicUniversity/rukopys-leaderboard --repo-type space --local-dir rukopys-leaderboard

Example: You open the Space's page to see how handwriting OCR systems did on the Ukrainian test pages. The page shows the results it reads from the results dataset. When a new run is merged there, it appears on the page on its own.

For newcomers and experts: OCR means a program turns a picture of writing into typed text. Here the writing is handwritten Ukrainian, and the page shows how different systems do on the same test pages. Start by opening the Space's page in your browser and looking at the results.

Under the hood, the Space is a single static HTML page. It reads its results from a separate dataset, so adding a run means changing that dataset and not the page. The evaluation code lives in a separate project named rukopys-eval, and the test pages come from the RUKOPYS dataset.

Test pages: RUKOPYS test split
Scoring: evaluation code in rukopys-eval
A run is merged into the results dataset
The static page reads that dataset
You see the updated results

Good to know: The page covers one narrow job, and the source names few details beyond that.

• License: cc-by-4, as the Space's license line states.
• Cost to run: none named.
• Data it reaches: the results dataset on Hugging Face, which the page reads. The scored pages come from the RUKOPYS test split.
• Not for: anything beyond whole-page OCR of Ukrainian handwriting on the public test split.

On Hugging Face, 13 November 2026

• Likes: 11
• Built with: static
• License: cc-by-4.0

Links

• Space: https://huggingface.co/spaces/UkrainianCatholicUniversity/rukopys-leaderboard
• Files: https://huggingface.co/spaces/UkrainianCatholicUniversity/rukopys-leaderboard/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.