Bram Vanroy
Researcher at Hugging Face
Belgium
Summary
🤩
Rockstar🎓
Top SchoolBram Vanroy is an NLP researcher and machine translation evaluation specialist with 11 years of experience bridging academic research and open-source engineering. He holds a PhD in Translation Studies and advanced degrees in AI and computational linguistics, and currently works on computational creativity projects at KU Leuven while also researching language technologies at the Dutch Language Institute. As a long-time Hugging Face core contributor, Bram has helped harden tokenization and training workflows in transformers and expanded evaluation tooling by adding metrics like TER and ChrF variants to the evaluate library. He contributes practical clarity to major open-source projects—improving documentation in spaCy and fixing subtle bugs that improve usability for thousands of users. Bram combines grant-winning academic work (building the MATEO MT evaluation platform) with hands-on coding and technical writing, making him as comfortable with reproducible research as with shipping library fixes. Colleagues rely on him for meticulous evaluation methodology and pragmatic improvements that often go unnoticed but materially raise tool quality.
11 years of coding experience
6 years of employment as a software developer
Master of Science (MSc) Artificial Intelligence, Master of Science (MSc) Artificial Intelligence at KU LeuvenBachelor of Arts (BA) Linguistics and literature English and Dutch, Bachelor of Arts (BA) Linguistics and literature English and Dutch at HUB-KUBrusselDoctor of Philosophy - PhD Translation Studies, Doctor of Philosophy - PhD Translation Studies at Ghent University
Stackoverflow
27,595reputation
11.3mreached
663answers
423questions
html
top-1%
checkbox
top-5%
css-animations
top-5%
css-transforms
top-5%
nlp
top-5%
javascript
top-1%
Github Skills (34)
tokenize10
pytorch10
javascript10
language-model10
jquery10
python10
evaluation10
css10
machine-learning10
huggingface-transformers10
tokenizer10
deep-learning10
spacy10
tensorflow10
trainings10
Programming languages (17)
Github contributions (5)
huggingface/transformers
Sep 2019 - Sep 2022
🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
Role in this project:
ML Engineer
ML EngineerContributions:13 reviews, 31 commits, 46 PRs in 2 years 11 months
Contributions summary:Bram primarily contributed to the Hugging Face Transformers library by addressing issues related to tokenization and model training. They implemented checks and made modifications to ensure correct behavior when handling padding tokens, especially for specific tokenizer types. Furthermore, they made changes to CLI tools by adding functionalities, refactoring and enhancing utility. The user also added parameters for cached_path and pretrained models.
pythonbertspeech-recognitionstate-of-the-artflax
huggingface/evaluate
Nov 2021 - Dec 2022
🤗 Evaluate: A library for easily evaluating machine learning models and datasets.
Role in this project:ML Engineer
ML EngineerContributions:10 reviews, 14 commits, 8 PRs in 1 year 1 month
Contributions summary:Bram contributed to the `evaluate` library by integrating and refining machine-learning evaluation metrics. Their work involved adding new metrics like TER, ChrF(++), CharacTER, and CharCut, leveraging the `sacrebleu` and other libraries. They also addressed bugs and improved the code base by clarifying error messages and fixing requirements. Their contributions focused on expanding the library's capabilities for evaluating machine learning models.
evaluateevaluation-frameworkdata-scienceevaluationmachine-learning
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.