Reinforcement Learning

This tutorial will present the basics of the Gymnasium and Stable-Baselines3 (SB3) libraries in order to apply reinforcement learning in practice. The session will cover the basics of how to create a custom task and solve it using algorithms from …

This dissertation makes several contributions to the training of reinforcement learning agents directly on real robots. It introduces a reliable software suite and a new exploration strategy to replace the standard step-based one. The thesis also …

A presentation on recent advances in RL, in terms of algorithms, software, and simulators.

Invited talk while visiting the INRIA Willow team in Paris.

Invited talk for the Soccer Robots workshop at Humanoids conference 2024

Talk at the Reinforcement Learning for Autonomous Accelerators workshop (RL4AA). The idea is to walk through the different steps of RL experimentation (task design, choosing the right algorithm, implementing safety layers) and also provide practical advice on how to run experiments and troubleshoot common problems.

Talk at the CDC 2023 Workshop on Benchmarking, Reproducibility, and Open-Source Code in Controls, about the lessons learned while developping Stable-Baselines3 to have reliable implementations and reproducible experiments.

From tabular Q-learning to Deep Q-Network (DQN)

Proof of concept version of Stable-Baselines3 in Jax.