Younes Belkada
Senior AI Engineer at Hugging Face
France
Summary
🤩
Rockstar🎓
Top SchoolYounes Belkada is a Senior AI Engineer based in France with six years of hands-on experience building and optimizing large-scale transformer and diffusion models. He has strong industry pedigree from Hugging Face—where he contributed to flagship projects like transformers, diffusers, accelerate, and PEFT—and currently works at the Technology Innovation Institute. His contributions span practical ML engineering (LoRA/PEFT integration, quantized model support, BetterTransformer optimizations) and reproducible developer experiences (notebooks and training script refactors), reflecting both research rigor and production sensibility. Younes combines deep academic training from EPFL and ENS Paris-Saclay with applied work across image segmentation, LLM fine-tuning, and model serialization. Notably, he has modernized critical Hugging Face tooling to support newer transformers and low-bit workflows, enabling more efficient training and inference. Colleagues describe him as a pragmatic problem-solver who bridges cutting-edge research and robust engineering.
6 years of coding experience
3 years of employment as a software developer
Master's degree Mathematics Computer Vision and Machine Learning, Master's degree Mathematics Computer Vision and Machine Learning at ENS Paris-SaclayFrench Baccalaureate Scientific, French Baccalaureate Scientific at Alexander Dumas International SchoolEngineering Curriculum Preparatory Class Engineering, Engineering Curriculum Preparatory Class Engineering at Pierre and Marie Curie UniversityEngineering Degree Applied Mathematics and Computer Science, Engineering Degree Applied Mathematics and Computer Science at Polytech SorbonneMaster's degree Data Science, Master's degree Data Science at EPFL
French, Arabic, Japanese, English
Github Skills (35)
transformers10
pytorch10
language-model10
python10
model-driven10
testing10
diffusion-models10
machine-learning10
stable-diffusion10
inference10
bart10
reinforcement-learning10
implement10
model-building10
data-loading10
Programming languages (9)
Github contributions (5)
huggingface/trl
Dec 2022 - Jan 2023
Train transformer language models with reinforcement learning.
Role in this project:
ML Engineer
ML EngineerContributions:28 releases, 837 reviews, 40 commits in 1 month
Contributions summary:Younes refactored and updated code related to training transformer language models with reinforcement learning. They modified existing scripts to work with newer versions of the `transformers` library, and they refactored the code, including moving notebooks to a dedicated directory. The user also updated the project's dependencies, indicating involvement in the project's machine learning aspects, and ensuring the code's compatibility with the latest libraries.
nlptransformer-architecturelanguage-modelfinetuninglanguage-modeling
huggingface/peft
Jan 2023 - Mar 2023
🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:1 release, 629 reviews, 19 commits in 1 month
Contributions summary:Younes's commits focused on implementing support for the `from_pretrained` method for both configuration and LoRA models within the PEFT library. They added support for loading and saving models, enabling the integration of pre-trained configurations and LoRA weights. Furthermore, tests were added to ensure the correctness of the code and its features, including a method for model serialization. The commits also provided examples and tests for using the library and its related models within the broader scope of Parameter-Efficient Fine-Tuning (PEFT).
parametermachine-learningfineparameter-efficientfine-tuning
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.