<linkrel="index"title="Index"href="../../genindex/"/><linkrel="search"title="Search"href="../../search/"/><linkrel="next"title="Solving Blackjack with Q-Learning"href="../../tutorials/blackjack_tutorial/"/><linkrel="prev"title="Zaxxon"href="../atari/zaxxon/"/>
<liclass="toctree-l1"><aclass="reference external"href="https://github.com/Farama-Foundation/Gymnasium/blob/main/docs/README.md">Contribute to the Docs</a></li>
<aclass="muted-link"href="https://github.com/Farama-Foundation/Gymnasium/edit/main/docs/environments/third_party_environments.md"title="Edit this page">
<p>Many environments that comply with the Gymnasium API are now maintained under the Farama Foundation’s <aclass="reference external"href="https://farama.org/projects">projects</a>, along with Gymnasium itself. These include many of the most popular environments using the Gymnasium API, and we encourage you to check them out. This page exclusively lists interesting third party environments that are not part of the Farama Foundation.</p>
<h3><aclass="reference external"href="https://github.com/Talendar/flappy-bird-gym"> flappy-bird-gym: A Flappy Bird environment for Gym</a><aclass="headerlink"href="#flappy-bird-gym-a-flappy-bird-environment-for-gym"title="Permalink to this heading">#</a></h3>
<p>A simple environment for single-agent reinforcement learning algorithms on a clone of <aclass="reference external"href="https://en.wikipedia.org/wiki/Flappy_Bird">Flappy Bird</a>, the hugely popular arcade-style mobile game. Both state and pixel observation environments are available.</p>
<h3><aclass="reference external"href="https://gymnasium.derkgame.com"> gym-derk: GPU accelerated MOBA environment</a><aclass="headerlink"href="#gym-derk-gpu-accelerated-moba-environment"title="Permalink to this heading">#</a></h3>
<p>This is a 3v3 MOBA environment where you train creatures to fight each other. It runs entirely on the GPU so you can easily have hundreds of instances running in parallel. There are around 15 items for the creatures, 60 “senses”, 5 actions, and roughly 23 tweakable rewards. It’s also possible to benchmark an agent against other agents online. It’s available for free for training for personal use, and otherwise costs money; see licensing details on the website</p>
<h3><aclass="reference external"href="https://github.com/hardmaru/slimevolleygym"> SlimeVolleyGym: A simple environment for single and multi-agent reinforcement learning</a><aclass="headerlink"href="#slimevolleygym-a-simple-environment-for-single-and-multi-agent-reinforcement-learning"title="Permalink to this heading">#</a></h3>
<p>A simple environment for benchmarking single and multi-agent reinforcement learning algorithms on a clone of Slime Volleyball game. The only dependencies are gym and NumPy. Both state and pixel observation environments are available. The motivation of this environment is to easily enable trained agents to play against each other, and also facilitate the training of agents directly in a multi-agent setting, thus adding an extra dimension for evaluating an agent’s performance.</p>
</section>
<sectionid="stable-retro">
<h3><aclass="reference external"href="https://github.com/MatPoliquin/stable-retro"> stable-retro</a><aclass="headerlink"href="#stable-retro"title="Permalink to this heading">#</a></h3>
<p>Supported fork of gym-retro with additional games, states, scenarios, etc. Open to PRs of additional games, features, and platforms since gym-retro is no longer maintained</p>
</section>
<sectionid="unity-ml-agents">
<h3><aclass="reference external"href="https://github.com/Unity-Technologies/ml-agents"> Unity ML Agents</a><aclass="headerlink"href="#unity-ml-agents"title="Permalink to this heading">#</a></h3>
<p>Gym wrappers for arbitrary and premade environments with the Unity game engine.</p>
<h3><aclass="reference external"href="https://github.com/qlan3/gym-games"> gym-games</a><aclass="headerlink"href="#gym-games"title="Permalink to this heading">#</a></h3>
<p>Gym implementations of the MinAtar games, various PyGame Learning Environment games, and various custom exploration games</p>
<h3><aclass="reference external"href="https://github.com/222464/PGE"> PGE: Parallel Game Engine</a><aclass="headerlink"href="#pge-parallel-game-engine"title="Permalink to this heading">#</a></h3>
<p>PGE is a FOSS 3D engine for AI simulations and can interoperate with the Gym. Contains environments with modern 3D graphics, and uses Bullet for physics.</p>
<h3><aclass="reference external"href="https://github.com/wil3/gymfc/">GymFC: A flight control tuning and training framework</a><aclass="headerlink"href="#gymfc-a-flight-control-tuning-and-training-framework"title="Permalink to this heading">#</a></h3>
<p>GymFC is a modular framework for synthesizing neuro-flight controllers. The architecture integrates digital twinning concepts to provide a seamless transfer of trained policies to hardware. The environment has been used to generate policies for the world’s first open-source neural network flight control firmware <aclass="reference external"href="https://github.com/wil3/neuroflight">Neuroflight</a>.</p>
<h3><aclass="reference external"href="https://github.com/erlerobot/gym-gazebo/">gym-gazebo</a><aclass="headerlink"href="#gym-gazebo"title="Permalink to this heading">#</a></h3>
<p>gym-gazebo presents an extension of the initial Gym for robotics using ROS and Gazebo, an advanced 3D modeling and
<h3><aclass="reference external"href="https://github.com/osannolik/gym-goddard">gym-goddard: Goddard’s Rocket Problem</a><aclass="headerlink"href="#gym-goddard-goddard-s-rocket-problem"title="Permalink to this heading">#</a></h3>
<p>An environment for simulating the classical optimal control problem where the thrust of a vertically ascending rocket shall be determined such that it reaches the maximum possible altitude while being subject to varying aerodynamic drag, gravity, and mass.</p>
<h3><aclass="reference external"href="https://github.com/Wandercraft/jiminy">gym-jiminy: training Robots in Jiminy</a><aclass="headerlink"href="#gym-jiminy-training-robots-in-jiminy"title="Permalink to this heading">#</a></h3>
<p>gym-jiminy presents an extension of the initial Gym for robotics using Jiminy, an extremely fast and light-weight simulator for poly-articulated systems using Pinocchio for physics evaluation and Meshcat for web-based 3D rendering.</p>
<h3><aclass="reference external"href="https://github.com/JacopoPan/gym-pybullet-drones">gym-pybullet-drones</a><aclass="headerlink"href="#gym-pybullet-drones"title="Permalink to this heading">#</a></h3>
<p>A simple environment using <aclass="reference external"href="https://github.com/bulletphysics/bullet3">PyBullet</a> to simulate the dynamics of a <aclass="reference external"href="https://www.bitcraze.io/documentation/hardware/crazyflie_2_1/crazyflie_2_1-datasheet.pdf">Bitcraze Crazyflie 2.x</a> nanoquadrotor.</p>
<h3><aclass="reference external"href="https://github.com/dimikout3/MarsExplorer">MarsExplorer</a><aclass="headerlink"href="#marsexplorer"title="Permalink to this heading">#</a></h3>
<p>Mars Explorer is a Ggym compatible environment designed and developed as an initial endeavor to bridge the gap between powerful Deep Reinforcement Learning methodologies and the problem of exploration/coverage of an unknown terrain.</p>
<h3><aclass="reference external"href="https://github.com/qgallouedec/panda-gym/">panda-gym </a><aclass="headerlink"href="#panda-gym"title="Permalink to this heading">#</a></h3>
<h3><aclass="reference external"href="https://github.com/jr-robotics/robo-gym">robo-gym</a><aclass="headerlink"href="#robo-gym"title="Permalink to this heading">#</a></h3>
<p>robo-gym provides a collection of reinforcement learning environments involving robotic tasks applicable in both simulation and real-world robotics.</p>
<h3><aclass="reference external"href="https://github.com/offworld-projects/offworld-gym">Offworld-gym</a><aclass="headerlink"href="#offworld-gym"title="Permalink to this heading">#</a></h3>
<h3><aclass="reference external"href="https://github.com/stanfordnmbl/osim-rl">osim-rl</a><aclass="headerlink"href="#osim-rl"title="Permalink to this heading">#</a></h3>
<p>Musculoskeletal Models in OpenSim. A human musculoskeletal model and a physics-based simulation environment where you can synthesize physically and physiologically accurate motion. One of the environments built in this framework is a competition environment for a NIPS 2017 challenge.</p>
</section>
<sectionid="safe-control-gym">
<h3><aclass="reference external"href="https://github.com/utiasDSL/safe-control-gym">safe-control-gym</a><aclass="headerlink"href="#safe-control-gym"title="Permalink to this heading">#</a></h3>
<p>PyBullet based CartPole and Quadrotor environments—with <aclass="reference external"href="https://web.casadi.org">CasADi</a> (symbolic) <em>a priori</em> dynamics and constraints—for learning-based control and model-based reinforcement learning.</p>
</section>
<sectionid="racecar-gym">
<h3><aclass="reference external"href="https://github.com/axelbr/racecar_gym/">racecar_gym</a><aclass="headerlink"href="#racecar-gym"title="Permalink to this heading">#</a></h3>
<p>A gym environment for a miniature racecar using the pybullet physics engine.</p>
</section>
<sectionid="jiminy">
<h3><aclass="reference external"href="https://github.com/duburcqa/jiminy/">jiminy</a><aclass="headerlink"href="#jiminy"title="Permalink to this heading">#</a></h3>
<p>Jiminy: a fast and portable Python/C++ simulator of poly-articulated systems with Gym interface for reinforcement learning</p>
</section>
<sectionid="gym-softrobot">
<h3><aclass="reference external"href="https://github.com/skim0119/gym-softrobot/">gym-softrobot</a><aclass="headerlink"href="#gym-softrobot"title="Permalink to this heading">#</a></h3>
<p>Softrobotics environment package for Gym</p>
</section>
<sectionid="ostrichrl">
<h3><aclass="reference external"href="https://github.com/vittorione94/ostrichrl/">ostrichrl</a><aclass="headerlink"href="#ostrichrl"title="Permalink to this heading">#</a></h3>
<p>This is the repository accompanying the paper <aclass="reference external"href="https://arxiv.org/abs/2112.06061">OstrichRL: A Musculoskeletal Ostrich Simulation to Study Bio-mechanical Locomotion</a>.</p>
</section>
<sectionid="quadruped-gym">
<h3><aclass="reference external"href="https://github.com/dtch1997/quadruped-gym/">quadruped-gym</a><aclass="headerlink"href="#quadruped-gym"title="Permalink to this heading">#</a></h3>
<p>A Gym environment for the training of legged robots</p>
</section>
<sectionid="evogym">
<h3><aclass="reference external"href="https://github.com/EvolutionGym/evogym/">evogym</a><aclass="headerlink"href="#evogym"title="Permalink to this heading">#</a></h3>
<p>A large-scale benchmark for co-optimizing the design and control of soft robots, as seen in NeurIPS 2021.</p>
</section>
<sectionid="igibson">
<h3><aclass="reference external"href="https://github.com/StanfordVL/iGibson/">iGibson</a><aclass="headerlink"href="#igibson"title="Permalink to this heading">#</a></h3>
<p>A Simulation Environment to train Robots in Large Realistic Interactive Scenes</p>
</section>
<sectionid="dexteroushands">
<h3><aclass="reference external"href="https://github.com/PKU-MARL/DexterousHands/">DexterousHands</a><aclass="headerlink"href="#dexteroushands"title="Permalink to this heading">#</a></h3>
<p>This is a library that provides dual dexterous hand manipulation tasks through Isaac Gym</p>
</section>
<sectionid="omniisaacgymenvs">
<h3><aclass="reference external"href="https://github.com/NVIDIA-Omniverse/OmniIsaacGymEnvs/">OmniIsaacGymEnvs</a><aclass="headerlink"href="#omniisaacgymenvs"title="Permalink to this heading">#</a></h3>
<p>Reinforcement Learning Environments for Omniverse Isaac Gym</p>
</section>
<sectionid="spacerobotenv">
<h3><aclass="reference external"href="https://github.com/Tsinghua-Space-Robot-Learning-Group/SpaceRobotEnv/">SpaceRobotEnv</a><aclass="headerlink"href="#spacerobotenv"title="Permalink to this heading">#</a></h3>
<p>A gym environment designed for free-floating space robot control based on the MuJoCo platform.</p>
</section>
<sectionid="gym-line-follower">
<h3><aclass="reference external"href="https://github.com/nplan/gym-line-follower/">gym-line-follower</a><aclass="headerlink"href="#gym-line-follower"title="Permalink to this heading">#</a></h3>
<p>Line follower robot simulator environment for Open AI Gym.</p>
<h2>Autonomous Driving and Traffic Control Environments<aclass="headerlink"href="#autonomous-driving-and-traffic-control-environments"title="Permalink to this heading">#</a></h2>
<sectionid="gym-carla">
<h3><aclass="reference external"href="https://github.com/cjy1992/gym-carla"> gym-carla</a><aclass="headerlink"href="#gym-carla"title="Permalink to this heading">#</a></h3>
<p>gym-carla provides a gym wrapper for the <aclass="reference external"href="http://carla.org/">CARLA simulator</a>, which is a realistic 3D simulator for autonomous driving research. The environment includes a virtual city with several surrounding vehicles running around. Multiple sources of observations are provided for the ego vehicle, such as front-view camera image, lidar point cloud image, and bird-eye view semantic mask. Several applications have been developed based on this wrapper, such as deep reinforcement learning for end-to-end autonomous driving.</p>
</section>
<sectionid="gym-duckietown">
<h3><aclass="reference external"href="https://github.com/duckietown/gym-duckietown"> gym-duckietown</a><aclass="headerlink"href="#gym-duckietown"title="Permalink to this heading">#</a></h3>
<p>A lane-following simulator built for the <aclass="reference external"href="http://duckietown.org/">Duckietown</a> project (small-scale self-driving car course).</p>
</section>
<sectionid="gym-electric-motor">
<h3><aclass="reference external"href="https://github.com/upb-lea/gym-electric-motor"> gym-electric-motor</a><aclass="headerlink"href="#gym-electric-motor"title="Permalink to this heading">#</a></h3>
<p>An environment for simulating a wide variety of electric drives taking into account different types of electric motors and converters. Control schemes can be continuous, yielding a voltage duty cycle, or discrete, determining converter switching states directly.</p>
</section>
<sectionid="highway-env">
<h3><aclass="reference external"href="https://github.com/eleurent/highway-env"> highway-env</a><aclass="headerlink"href="#highway-env"title="Permalink to this heading">#</a></h3>
<p>An environment for behavioral planning in autonomous driving, with an emphasis on high-level perception and decision rather than low-level sensing and control. The difficulty of the task lies in understanding the social interactions with other drivers, whose behaviors are uncertain. Several scenes are proposed, such as highway, merge, intersection and roundabout.</p>
</section>
<sectionid="longicontrol">
<h3><aclass="reference external"href="https://github.com/dynamik1703/gym_longicontrol"> LongiControl</a><aclass="headerlink"href="#longicontrol"title="Permalink to this heading">#</a></h3>
<p>An environment for the stochastic longitudinal control of an electric vehicle. It is intended to be a descriptive and comprehensible example of a continuous real-world problem within the field of autonomous driving.</p>
</section>
<sectionid="sumo-rl">
<h3><aclass="reference external"href="https://github.com/LucasAlegre/sumo-rl"> sumo-rl</a><aclass="headerlink"href="#sumo-rl"title="Permalink to this heading">#</a></h3>
<p>Gym wrapper for various environments in the Sumo traffic simulator</p>
</section>
<sectionid="commonroad-rl">
<h3><aclass="reference external"href="https://commonroad.in.tum.de/tools/commonroad-rl"> CommonRoad-RL</a><aclass="headerlink"href="#commonroad-rl"title="Permalink to this heading">#</a></h3>
<p>A Gym for solving motion planning problems for various traffic scenarios compatible with <aclass="reference external"href="https://commonroad.in.tum.de/scenarios">CommonRoad benchmarks</a>, which provides configurable rewards, action spaces, and observation spaces.</p>
<h3><aclass="reference external"href="https://github.com/trackmania-rl/tmrl/">tmrl</a><aclass="headerlink"href="#tmrl"title="Permalink to this heading">#</a></h3>
<p>TrackMania 2020 through RL</p>
</section>
<sectionid="racing-dreamer">
<h3><aclass="reference external"href="https://github.com/CPS-TUWien/racing_dreamer/">racing_dreamer</a><aclass="headerlink"href="#racing-dreamer"title="Permalink to this heading">#</a></h3>
<p>Latent Imagination Facilitates Zero-Shot Transfer in Autonomous Racing</p>
</section>
<sectionid="l2r">
<h3><aclass="reference external"href="https://github.com/learn-to-race/l2r/">l2r</a><aclass="headerlink"href="#l2r"title="Permalink to this heading">#</a></h3>
<p>An open-source reinforcement learning environment for autonomous racing.</p>
</section>
<sectionid="gym-torcs">
<h3><aclass="reference external"href="https://github.com/ugo-nama-kun/gym_torcs/">gym_torcs</a><aclass="headerlink"href="#gym-torcs"title="Permalink to this heading">#</a></h3>
<p>Gym-TORCS is the reinforcement learning (RL) environment in TORCS domain with gym-like interface. TORCS is the open-source realistic car racing simulator recently used as an RL benchmark task in several AI studies.</p>
<h2>Recommendation System Environments<aclass="headerlink"href="#recommendation-system-environments"title="Permalink to this heading">#</a></h2>
<sectionid="gym-adserve">
<h3><aclass="reference external"href="https://github.com/falox/gym-adserver"> gym-adserve</a><aclass="headerlink"href="#gym-adserve"title="Permalink to this heading">#</a></h3>
<p>An environment that implements a typical <aclass="reference external"href="https://en.wikipedia.org/wiki/Multi-armed_bandit">multi-armed bandit scenario</a> where an <aclass="reference external"href="https://en.wikipedia.org/wiki/Ad_serving">ad server</a> must select the best advertisement to be displayed in a web page. Some example agents included: Random, epsilon-Greedy, Softmax, and UCB1.</p>
</section>
<sectionid="gym-recsys">
<h3><aclass="reference external"href="https://github.com/zuoxingdong/gym-recsys"> gym-recsys</a><aclass="headerlink"href="#gym-recsys"title="Permalink to this heading">#</a></h3>
<p>This package describes an Gym interface for creating a simulation environment of reinforcement learning-based recommender systems (RL-RecSys). The design strives for simple and flexible APIs to support novel research.</p>
</section>
<sectionid="virtualtaobao">
<h3><aclass="reference external"href="https://github.com/eyounx/VirtualTaobao/"> VirtualTaobao</a><aclass="headerlink"href="#virtualtaobao"title="Permalink to this heading">#</a></h3>
<p>An environment for online recommendations, where customers are learned from Taobao.com, one of the world’s largest e-commerce platforms.</p>
</section>
</section>
<sectionid="industrial-process-environments">
<h2>Industrial Process Environments<aclass="headerlink"href="#industrial-process-environments"title="Permalink to this heading">#</a></h2>
<sectionid="gym-inventory">
<h3><aclass="reference external"href="https://github.com/paulhendricks/gym-inventory"> gym-inventory</a><aclass="headerlink"href="#gym-inventory"title="Permalink to this heading">#</a></h3>
<p>gym-inventory is a single-agent domain featuring discrete state and action spaces that an AI agent might encounter in inventory control problems.</p>
</section>
<sectionid="openmodelica-microgrid-gym">
<h3><aclass="reference external"href="https://github.com/upb-lea/openmodelica-microgrid-gym"> openmodelica-microgrid-gym</a><aclass="headerlink"href="#openmodelica-microgrid-gym"title="Permalink to this heading">#</a></h3>
<p>The OpenModelica Microgrid Gym (OMG) package is a software toolbox for the simulation and control optimization of microgrids based on energy conversion by power electronic converters.</p>
</section>
<sectionid="mobile-env">
<h3><aclass="reference external"href="https://github.com/stefanbschneider/mobile-env/">mobile-env</a><aclass="headerlink"href="#mobile-env"title="Permalink to this heading">#</a></h3>
<p>An open, minimalist Gym environment for autonomous coordination in wireless mobile networks.</p>
</section>
<sectionid="pyelastica">
<h3><aclass="reference external"href="https://github.com/GazzolaLab/PyElastica/">PyElastica</a><aclass="headerlink"href="#pyelastica"title="Permalink to this heading">#</a></h3>
<p>Python implementation of Elastica, open-source software for the simulation of assemblies of slender, one-dimensional structures using Cosserat Rod theory.</p>
</section>
</section>
<sectionid="financial-environments">
<h2>Financial Environments<aclass="headerlink"href="#financial-environments"title="Permalink to this heading">#</a></h2>
<sectionid="gym-anytrading">
<h3><aclass="reference external"href="https://github.com/AminHP/gym-anytrading"> gym-anytrading</a><aclass="headerlink"href="#gym-anytrading"title="Permalink to this heading">#</a></h3>
<p>AnyTrading is a collection of Gym environments for reinforcement learning-based trading algorithms with a great focus on simplicity, flexibility, and comprehensiveness.</p>
</section>
<sectionid="gym-mtsim">
<h3><aclass="reference external"href="https://github.com/AminHP/gym-mtsim"> gym-mtsim</a><aclass="headerlink"href="#gym-mtsim"title="Permalink to this heading">#</a></h3>
<h2>Other Environments<aclass="headerlink"href="#other-environments"title="Permalink to this heading">#</a></h2>
<sectionid="carl">
<h3><aclass="reference external"href="https://github.com/automl/CARL"> CARL</a><aclass="headerlink"href="#carl"title="Permalink to this heading">#</a></h3>
<p>Configurable reinforcement learning environments for testing generalization, e.g. CartPole with variable pole lengths or Brax robots with different ground frictions.</p>
</section>
<sectionid="compilergym">
<h3><aclass="reference external"href="https://github.com/facebookresearch/CompilerGym"> CompilerGym</a><aclass="headerlink"href="#compilergym"title="Permalink to this heading">#</a></h3>
<p>Reinforcement learning environments for compiler optimization tasks, such as LLVM phase ordering, GCC flag tuning, and CUDA loop nest code generation.</p>
</section>
<sectionid="dacbench">
<h3><aclass="reference external"href="https://github.com/automl/DACBench"> DACBench</a><aclass="headerlink"href="#dacbench"title="Permalink to this heading">#</a></h3>
<p>Environments for hyperparameter configuration using RL. Includes cheap surrogate benchmarks as well as real-world algorithms from e.g. AI Planning, Evolutionary Computation and Deep Learning.</p>
</section>
<sectionid="gym-autokey">
<h3><aclass="reference external"href="https://github.com/Flunzmas/gym-autokey"> gym-autokey</a><aclass="headerlink"href="#gym-autokey"title="Permalink to this heading">#</a></h3>
<p>An environment for automated rule-based deductive program verification in the KeY verification system.</p>
</section>
<sectionid="gym-cellular-automata">
<h3><aclass="reference external"href="https://github.com/elbecerrasoto/gym-cellular-automata"> gym-cellular-automata</a><aclass="headerlink"href="#gym-cellular-automata"title="Permalink to this heading">#</a></h3>
<p>Environments where the agent interacts with <em>Cellular Automata</em> by changing its cell states.</p>
</section>
<sectionid="gym-maze">
<h3><aclass="reference external"href="https://github.com/tuzzer/gym-maze/"> gym-maze</a><aclass="headerlink"href="#gym-maze"title="Permalink to this heading">#</a></h3>
<p>A simple 2D maze environment where an agent finds its way from the start position to the goal.</p>
<h3><aclass="reference external"href="https://github.com/erfanMhi/gym-riverswim"> gym-riverswim</a><aclass="headerlink"href="#gym-riverswim"title="Permalink to this heading">#</a></h3>
<p>A simple environment for benchmarking reinforcement learning exploration techniques in a simplified setting. Hard exploration.</p>
</section>
<sectionid="gym-sokoban">
<h3><aclass="reference external"href="https://github.com/mpSchrader/gym-sokoban"> gym-sokoban</a><aclass="headerlink"href="#gym-sokoban"title="Permalink to this heading">#</a></h3>
<p>2D Transportation Puzzles. The environment consists of transportation puzzles in which the player’s goal is to push all boxes to the warehouse’s storage locations. The advantage of the environment is that it generates a new random level every time it is initialized or reset, which prevents overfitting to predefined levels.</p>
</section>
<sectionid="math-prog-synth-env">
<h3><aclass="reference external"href="https://github.com/JohnnyYeeee/math_prog_synth_env"> math-prog-synth-env</a><aclass="headerlink"href="#math-prog-synth-env"title="Permalink to this heading">#</a></h3>
<p>In our paper “A Reinforcement Learning Environment for Mathematical Reasoning via Program Synthesis” we convert the DeepMind Mathematics Dataset into an RL environment based on program synthesis.https://arxiv.org/abs/2107.07373</p>
</section>
<sectionid="nasgym">
<h3><aclass="reference external"href="https://github.com/gomerudo/nas-env"> NASGym</a><aclass="headerlink"href="#nasgym"title="Permalink to this heading">#</a></h3>
<p>The environment is fully compatible with the OpenAI baselines and exposes a NAS environment following the Neural Structure Code of <aclass="reference external"href="https://arxiv.org/abs/1808.05584">BlockQNN: Efficient Block-wise Neural Network Architecture Generation</a>. Under this setting, a Neural Network (i.e. the state for the reinforcement learning agent) is modeled as a list of NSCs, an action is the addition of a layer to the network, and the reward is the accuracy after the early-stop training. The datasets considered so far are the CIFAR-10 dataset (available by default) and the meta-dataset (has to be manually downloaded as specified in <aclass="reference external"href="https://github.com/gomerudo/meta-dataset">this repository</a>).</p>
<h3><aclass="reference external"href="https://github.com/rajcscw/nlp-gym"> NLPGym: A toolkit to develop RL agents to solve NLP tasks</a><aclass="headerlink"href="#nlpgym-a-toolkit-to-develop-rl-agents-to-solve-nlp-tasks"title="Permalink to this heading">#</a></h3>
<p><aclass="reference external"href="https://arxiv.org/pdf/2011.08272v1.pdf">NLPGym</a> provides interactive environments for standard NLP tasks such as sequence tagging, question answering, and sequence classification. Users can easily customize the tasks with their datasets, observations, features and reward functions.</p>
</section>
<sectionid="obstacle-tower">
<h3><aclass="reference external"href="https://github.com/Unity-Technologies/obstacle-tower-env"> Obstacle Tower</a><aclass="headerlink"href="#obstacle-tower"title="Permalink to this heading">#</a></h3>
<p>3D procedurally generated tower where you have to climb to the highest level possible</p>
</section>
<sectionid="qasgym">
<h3><aclass="reference external"href="https://github.com/qdevpsi3/quantum-arch-search"> QASGym</a><aclass="headerlink"href="#qasgym"title="Permalink to this heading">#</a></h3>
<p>This is a list of environments for quantum architecture search following the description in <aclass="reference external"href="https://arxiv.org/abs/2104.07715">Quantum Architecture Search via Deep Reinforcement Learning</a>. The agent designs the quantum circuit by taking actions in the environment. Each action corresponds to a gate applied on some wires. The goal is to build a circuit U that generates the target n-qubit quantum state that belongs to the environment and is hidden from the agent. The circuits are built using <aclass="reference external"href="https://quantumai.google/cirq">Google QuantumAI Cirq</a>.</p>
</section>
<sectionid="mo-gym">
<h3><aclass="reference external"href="https://github.com/LucasAlegre/mo-gym"> mo-gym</a><aclass="headerlink"href="#mo-gym"title="Permalink to this heading">#</a></h3>
<p>Multi-objective RL (MORL) gym environments, where the reward is a NumPy array of different (possibly conflicting) objectives.</p>
</section>
<sectionid="gym-saturation">
<h3><aclass="reference external"href="https://github.com/inpefess/gym-saturation">gym-saturation</a><aclass="headerlink"href="#gym-saturation"title="Permalink to this heading">#</a></h3>
<p>An environment for guiding automated theorem provers based on saturation algorithms (e.g. <aclass="reference external"href="https://github.com/vprover/vampire">Vampire</a>).</p>
</section>
<sectionid="shinrl">
<h3><aclass="reference external"href="https://github.com/omron-sinicx/ShinRL/">ShinRL</a><aclass="headerlink"href="#shinrl"title="Permalink to this heading">#</a></h3>
<p>ShinRL: A Library for Evaluating RL Algorithms from Theoretical and Practical Perspectives (Deep RL Workshop 2021)</p>
</section>
<sectionid="racing-rl">
<h3><aclass="reference external"href="https://github.com/luigiberducci/racing-rl/">racing-rl</a><aclass="headerlink"href="#racing-rl"title="Permalink to this heading">#</a></h3>
<h3><aclass="reference external"href="https://github.com/DoubleGremlin181/RubiksCubeGym"> RubiksCubeGym</a><aclass="headerlink"href="#rubikscubegym"title="Permalink to this heading">#</a></h3>
<p>The RubiksCubeGym package provides environments for twisty puzzles with multiple reward functions to help simulate the methods used by humans.</p>
<h3><aclass="reference external"href="https://github.com/EvolutionGym/evogym-design-tool/">evogym-design-tool</a><aclass="headerlink"href="#evogym-design-tool"title="Permalink to this heading">#</a></h3>
<p>Design tool for creating Evolution Gym environments.</p>
</section>
<sectionid="starship-landing-gym">
<h3><aclass="reference external"href="https://github.com/Armandpl/starship-landing-gym/">starship-landing-gym</a><aclass="headerlink"href="#starship-landing-gym"title="Permalink to this heading">#</a></h3>
<h3><aclass="reference external"href="https://github.com/chaosprint/RaveForce/">RaveForce</a><aclass="headerlink"href="#raveforce"title="Permalink to this heading">#</a></h3>
<h3><aclass="reference external"href="https://github.com/RobertTLange/gymnax/">gymnax</a><aclass="headerlink"href="#gymnax"title="Permalink to this heading">#</a></h3>
<li><aclass="reference internal"href="#slimevolleygym-a-simple-environment-for-single-and-multi-agent-reinforcement-learning"> SlimeVolleyGym: A simple environment for single and multi-agent reinforcement learning</a></li>
<li><aclass="reference internal"href="#gymfc-a-flight-control-tuning-and-training-framework">GymFC: A flight control tuning and training framework</a></li>
<li><aclass="reference internal"href="#autonomous-driving-and-traffic-control-environments">Autonomous Driving and Traffic Control Environments</a><ul>
<li><aclass="reference internal"href="#nlpgym-a-toolkit-to-develop-rl-agents-to-solve-nlp-tasks"> NLPGym: A toolkit to develop RL agents to solve NLP tasks</a></li>
