DLR-RM/rl-baselines3-zoo

Health v2 · maintenance only

Adoption not part of the score

2870 stars · 603 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

RL Baselines3 Zoo is a training framework for reinforcement learning agents built on Stable Baselines3. It provides CLI scripts for training, evaluating, hyperparameter tuning, and video recording, plus a collection of tuned hyperparameters and pre-trained agents for common environments.

Use cases

• train reinforcement learning agents on gym environments
• tune hyperparameters for RL algorithms
• benchmark different RL algorithms
• download pre-trained RL agents
• record videos of trained agents
• run hyperparameter search for SAC or PPO

When to choose

• you use Stable Baselines3 and want a ready-made training pipeline
• you need tuned hyperparameters for standard Gym/Gymnasium or PyBullet environments
• you want to benchmark RL algorithms consistently

When to avoid

• you need custom RL algorithms not in Stable Baselines3
• you want a GUI or notebook-first workflow
• you are not using PyTorch-based SB3

Facets

framework · maturity active

machine-learning reinforcement-learning benchmarking cli reinforcement-learning machine-learning robotics developer-tools python windows cli stable-baselines3 pytorch gymnasium pybullet hyperparameter-tuning pretrained-agents openai-gym linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="DLR-RM/rl-baselines3-zoo")