Regis Pierrard
Machine Learning Engineer at Hugging Face
France
Summary
🤩
Rockstar🎓
Top SchoolRegis Pierrard is a Machine Learning Engineer with a decade of experience building and optimizing ML systems from research prototypes to production deployments. Currently a core maintainer on Hugging Face's Optimum and Optimum Habana projects, he specializes in model optimization, ONNX export, and accelerator integration—work that directly improves deployment on platforms like Habana Gaudi and ONNX Runtime. His background includes a PhD in Explainable AI and publications on relational and fuzzy methods, which inform his emphasis on interpretable, efficient models for medical and embedded vision applications. He has hands-on expertise in PyTorch, TensorRT, graph optimizations and quantization, and has contributed to the flagship transformers library to enhance reproducibility and memory-efficient generation. Colleagues value him for bridging deep academic research with practical engineering, for example shipping trainer foundations and CI infrastructure that made hardware-specific tooling usable by the wider community.
10 years of coding experience
2 years of employment as a software developer
Master of Engineering - MEng, Computer Science, Master of Engineering - MEng, Computer Science at SupélecPreparation for national competitive entrance exams to leading French grandes écoles, Preparation for national competitive entrance exams to leading French grandes écoles at Lycée Louis de CormontaigneDoctor of Philosophy - PhD, Artificial Intelligence, Doctor of Philosophy - PhD, Artificial Intelligence at Université Paris-Saclay
French, English, German, Russian, Spanish
Github Skills (23)
transformers10
pytorch10
python10
package-management10
image-classification10
machine-learning10
onnx10
pipelining10
huggingface-transformers10
onnxruntime10
natural-language-processing10
pipe10
model-optimization10
pipeline10
computer-vision10
Programming languages (7)
Github contributions (5)
huggingface/optimum-habana
Mar 2022 - Jan 2023
Easy and lightning fast training of 🤗 Transformers on Habana Gaudi processor (HPU)
Role in this project:
Backend Developer
Backend DeveloperContributions:32 releases, 2012 reviews, 186 commits in 11 months
Contributions summary:Regis's primary contribution involved setting up the initial structure and core components for the `optimum-habana` library, specifically targeting the integration of Habana's Gaudi processor with Hugging Face's Transformers. This included creating necessary directories, adding a setup file to handle package configurations, integrating a GitHub workflow for code format checking, and adding a Makefile to manage the project's build and testing process. Furthermore, the user implemented the foundational elements for a training module, including trainer classes and configuration options.
transformersbertfine-tuninghabanahpu
huggingface/optimum
Apr 2022 - Dec 2022
🚀 Accelerate inference and training of 🤗 Transformers, Diffusers, TIMM and Sentence Transformers with easy to use hardware optimization tools
Role in this project:ML Engineer
ML EngineerContributions:7 releases, 344 reviews, 28 commits in 8 months
Contributions summary:Regis primarily contributed to the ONNX Runtime integration within the `optimum` repository, focusing on enhancing and expanding its capabilities for various machine learning tasks. Their commits included adding support for image classification pipelines, fixing bugs related to loading and utilizing models within the pipeline framework, and addressing version compatibility issues, especially with Hugging Face Hub and ONNX Runtime. Furthermore, the user implemented performance improvements through lazy imports and corrected documentation to align with the model functionalities, showing commitment to the usability and performance.
diffusershardwareinferencesentence-transformerstransformers
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.