Jean De Pomereu
United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolJean De Pomereu is a multidisciplinary polar historian and visual culture specialist with 13 years' experience researching the cultural, scientific and political histories of ice sheets at institutions including the Scott Polar Research Institute and the University of Cambridge as a Marie Skłodowska-Curie Research Fellow. Their work blends archival curation, editorial practice and exhibition-making—publishing and curating projects on early Antarctic photography and co-authoring "Antarctica: A History in 100 Objects"—while maintaining active studio practice as an exhibiting artist. Jean also brings practical technical contributions to open-source projects, having improved NLP data pipelines and vocabulary handling in Facebook Research’s pytext and localized Mozilla’s Common Voice into Ligurian, highlighting an unusual bridge between humanities scholarship and hands-on computational work. Based in the UK, they combine deep field expertise on polar science and climate history with a practiced eye for photographic and print techniques developed through two decades of editorial and printing work.
13 years of coding experience
4 years of employment as a software developer
Master of Philosophy - MPhil, Polar Studies, Master of Philosophy - MPhil, Polar Studies at University of CambridgeBachelor's degree, Art History, Bachelor's degree, Art History at The American University of ParisDoctor of Philosophy - PhD, Historical Geography, Doctor of Philosophy - PhD, Historical Geography at University of Exeter
French, English
Github Skills (8)
pytorch10
localization10
user-interface10
nlp10
translation10
python10
data-processing10
machine-learning9
Programming languages (21)
Github contributions (5)
facebookresearch/pytext
Sep 2019 - Oct 2020
A natural language modeling framework based on PyTorch
Role in this project:
ML Engineer
ML EngineerContributions:1 release, 16 commits, 14 PRs in 1 year 1 month
Contributions summary:Jean made several contributions focused on improving the `pytext` framework, which is built for natural language modeling. These changes involved modifying the data processing pipeline and vocabulary handling, which is key to NLP tasks. Specifically, the user exposed and made customizable special tokens like UNK, BOS, and EOS within the `VocabBuilder` and `Vocabulary` classes, and updated various tensorizer components.
pytorchnlpbertmachine-learningnatural-language
common-voice/common-voice
Mar 2019 - Jun 2020
Common Voice is part of Mozilla's initiative to help teach machines how real people speak.
Role in this project:Localization / Internationalization Specialist
Localization / Internationalization SpecialistContributions:71 commits in 1 year 2 months
Contributions summary:Jean's contributions center around updating the Ligurian (lij) localization for the Common Voice project. Their commits primarily involve modifications to `.ftl` files, which likely contain the language-specific text strings for the user interface. The user is responsible for translating and adapting the project's user interface to the Ligurian language. The commits demonstrate a focus on ensuring the project's usability for Ligurian speakers by adding new text and translating existing phrases.
internet-freedommachinesinitiativeopen-datavoice
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.