Stable Baselines3

Stable Baselines3 (SB3) is a collection of reliable reinforcement learning algorithm implementations built with PyTorch and is the next major version of Stable Baselines. It provides algorithms such as A2C, DDPG, DQN, HER, PPO, SAC, and TD3 in a unified structure, and the performance of each algorithm has been tested. It was built so that the research and industry communities can reproduce and improve results and use it as a baseline for comparing new methods with existing ones. It comes with the training framework RL Baselines3 Zoo, the experimental algorithms package SB3 Contrib, and the Jax version SBX, and it assumes basic knowledge of reinforcement learning.

• A2C, DDPG, DQN, HER, PPO, SAC, and TD3 algorithms
• Unified structure and common interface for all algorithms
• Performance test results provided for each algorithm
• TensorBoard support
• Custom environments and policies
• Vectorized environments (multiprocessing)
• Exporting models to ONNX, TensorFlow.js, TFLite, and more
• Integration with Weights & Biases, Hugging Face, and MLflow

Open source + paid cloud-managed plans

Open-source version can be self-hosted for free, managed cloud services are paid based on usage/seats

Estimate based on AI knowledge - please verify on the official site

• One of 45 AI tools in the MLOps & Training field.
• Connected to 5 other tools — #11 in its field.