Reinforcement Learning Python Packages

sglang

SGLang is a high-performance serving framework for large language models and multimodal models.

ray

Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.

wandb

The AI developer platform. Use Weights & Biases to train and fine-tune models, and manage models from experimentation to production.

gymnasium

A standard API for single-agent reinforcement learning environments, with popular reference environments and related utilities (formerly Gym)

unsloth

Unsloth Studio is a web UI for training and running open models like Gemma 4, Qwen3.6, DeepSeek, gpt-oss locally.

stable-baselines3

PyTorch version of Stable Baselines, reliable implementations of reinforcement learning algorithms.

unsloth-zoo

Unsloth Studio is a web UI for training and running open models like Gemma 4, Qwen3.6, DeepSeek, gpt-oss locally.

torchrl

A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.

torchtnt

A lightweight library for PyTorch training tools and utilities

pybullet

Bullet Physics SDK: real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning etc.

sglang-kernel

SGLang is a high-performance serving framework for large language models and multimodal models.

dm-control

Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.

sb3-contrib

Contrib package for Stable-Baselines3 - Experimental reinforcement learning (RL) code

shimmy

PettingZoo and Gymnasium bindings for popular reinforcement learning environments outside of Farama

pettingzoo

A standard API for multi-agent reinforcement learning environments, with popular reference environments and related utilities

open-spiel

OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.

sgl-kernel

SGLang is a high-performance serving framework for large language models and multimodal models.

mjlab

Isaac Lab API, powered by MuJoCo-Warp, for RL and robotics research

reasoning-gym

[NeurIPS 2025 Spotlight] Reasoning Environments for Reinforcement Learning with Verifiable Rewards

robosuite

robosuite: A Modular Simulation Framework and Benchmark for Robot Learning

mooncake-transfer-engine

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

fhlmi

Framework enabling modular interchange of language agents, environments, and optimizers

judgeval

The Continuous-Improvement Stack for Agents. Our environment data and evals power agent improvement and monitoring.

skrl

Modular Reinforcement Learning (RL) library (implemented in PyTorch, JAX, and NVIDIA Warp) with support for Gymnasium/Gym, NVIDIA Isaac Lab, MuJoCo Playground and other environments