Single-agent and Multi-agent Support

Train on-policy, off-policy, offline, multi-agent and contextual multi-armed bandits with unmatched speed and performance

Evolutionary Hyperparameter Optimization

Achieve automatic, optimal performance in a single training run through hyperparameter and neural network evolution

Distributed Training

Take full advantage of your entire compute stack for online and offline reinforcement learning with multi-GPU support

Hierarchical Skills

Solve complex problems by breaking down tasks into smaller, learnable sub-tasks with the AgileRL Skills wrapper

Framework Docs

GitHub