stable-baselines3 - Production-ready reinforcement learning algorithms library
PyTorch-based RL algorithms library for training agents and creating custom environments
Tags
Updated: 2026-06-30Capabilities
Typical Inputs
Typical Outputs
What this skill does
- train RL agents
- create custom environments
- implement callbacks
- save models
- load models
- evaluate agents
- record videos
- vectorize environments
- monitor training metrics
- optimize training workflows
Inputs
- Gymnasium environment
- algorithm type
- training hyperparameters
- callback definitions
- policy configuration
- learning rate schedule
Outputs
- trained model file
- evaluation metrics
- video recordings
- TensorBoard logs
- model checkpoints
Requirements
- Python environment
- PyTorch
- Gymnasium
- stable-baselines3 package