stable-baselines3

SUMMARY

What it does

This Agent Skill provides procedural guidance for using Stable Baselines3 (SB3), a PyTorch-based library for reinforcement learning. The skill covers training RL agents with algorithms like PPO, SAC, DQN, TD3, DDPG, and A2C; creating custom Gymnasium environments; using vectorized environments; implementing callbacks; model persistence; evaluation; and advanced features like learning rate schedules and TensorBoard integration. It includes scripts and reference documentation for common workflows. The skill is part of the Scientific Agent Skills collection by K-Dense-AI and is distributed under the MIT license.