Antonin Raffin
Researcher at German Aerospace Center (DLR)
Bavaria, Germany
Summary
🤩
Rockstar🎓
Top SchoolAntonin Raffin is a research engineer and roboticist with a decade of experience applying machine learning—especially reinforcement learning—to real-world robot systems. Based in Bavaria at the German Aerospace Center, he blends rigorous research (Dr. level studies at TUM) with hands-on open-source development, contributing core RL implementations to Stable-Baselines3 and training/tooling frameworks used across the community. His work spans algorithm design (TD3, TQC), reproducible training pipelines, and simulation integration (PyBullet, Gym) while also improving test infrastructure and documentation for broad adoption. Notably, he has extended classical robotics libraries (PythonRobotics) with path-planning primitives like Bezier curvature tools, reflecting a rare mix of control, simulation, and deep-RL expertise.
10 years of coding experience
1 year of employment as a software developer
Engineer’s Degree, Robotique et Systèmes Embarqués, Engineer’s Degree, Robotique et Systèmes Embarqués at ENSTA ParisTech - École Nationale Supérieure de Techniques AvancéesDr. rer. nat., Mechatronics, Robotics, and Automation Engineering, Dr. rer. nat., Mechatronics, Robotics, and Automation Engineering at Technical University of MunichClasses préparatoires Aux Lazaristes PCSI - PC*Master’s Degree, M2 Apprentissage, Information et Contenu - Machine Learning, Information and Content, Master’s Degree, M2 Apprentissage, Information et Contenu - Machine Learning, Information and Content at Université Paris-Saclay
French, English, German, Spanish
Github Skills (38)
algorithm10
simulation10
pytorch10
debug10
ppp10
robotics10
pytest10
python10
testing10
simulations10
gymnasium10
openai-gym10
machine-learning10
reinforcement-learning10
hyperparameter-optimization10
Programming languages (13)
Github contributions (5)
Stable-Baselines-Team/stable-baselines3-contrib
Sep 2020 - Jan 2023
Contrib package for Stable-Baselines3 - Experimental reinforcement learning (RL) code
Role in this project:
ML Engineer
ML EngineerContributions:17 releases, 156 reviews, 113 commits in 2 years 4 months
Contributions summary:Antonin contributed significantly to the implementation and maintenance of the TQC (Truncated Quantile Critics) algorithm within the Stable-Baselines3-Contrib repository. Their commits include the addition of core TQC components like policies, critic networks, and related scripts. The user was responsible for code cleanup, dependency updates, and improvements to test coverage, demonstrating a focus on integrating a new reinforcement learning algorithm into the project.
contribstable-baselinespybulletmahjonggsde
DLR-RM/stable-baselines3
Sep 2019 - Jan 2023
PyTorch version of Stable Baselines, reliable implementations of reinforcement learning algorithms.
Role in this project:Back-end Developer
Back-end DeveloperContributions:23 releases, 855 reviews, 675 commits in 3 years 5 months
Contributions summary:Antonin implemented and initialized TD3 (Twin Delayed Deep Deterministic Policy Gradients) algorithm with PyTorch for reinforcement learning. The implementation involved the creation of a dedicated TD3 class that inherits from the base RL model. This class includes methods for selecting actions, training the critic and actor, and storing the transitions in a replay buffer.
pythonstable-baselinesrobustnessgsdesde
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.