Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Open-source implementations of OpenAI Gym MuJoCo environments for use with the OpenAI Gym Reinforcement Learning Research Platform.
| Date | Stars |
|---|---|
| 2026-07-31 | 881 |
| 2026-08-04 | 881 |
| 2026-08-05 | 881 |
| 2026-08-06 | 881 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
PyBullet Gymperium
-----------
*PyBullet Gymperium is an open-source implementation of the OpenAI Gym MuJoCo environments for use with the OpenAI Gym Reinforcement Learning Research Platform in support of open research.*
OpenAI gym is currently one of the most widely used toolkit for developing and comparing reinforcement learning algorithms. Unfortunately, for several challenging continuous control environments it requires the user to install MuJoCo, a commercial physics engine which requires a license to run for longer than 30 days. Such a commercial barrier hinders open research, especially in the perspective that other appropriate physics engines exist. This repository provides alternative implementations of the original MuJoCo environments which can be used for free. The environments have been reimplemented using [BulletPhysics'](https://github.com/bulletphysics/bullet3) python wrapper pybullet, such that they seamlessly integrate into the OpenAI gym framework. In order to show the usability of the new environments, several RL agents from the [Tensorforce](https://github.com/reinforceio/tensorforce) Reinforcement Learning Library are configured to be trainable out of the box. To simplify research with the implemented environment, each environment is featured with pretrained agents which serve as unit tests for the implementations and could well serve as baselines for other purposes. <!--To further simplify the training of agents, a Trainer class was implemented which helps to capture commandline arguments in a unified fashion. The Trainer provides a set of standard arguments, but additional arguments can be defined by the agent and the environment to enable the researcher to provide special parameters to either one.-->
If you find our work useful in your research please consider citing as follows:
@misc {benelot2018,
author = {Benjamin Ellenberger},
title = {PyBullet Gymperium},
howpublished = {\url{ https://github.com/benelot/pybullet-gym}} ,
year = {2018--2019}
}
## State of implementations
Environment Name | Implemented | Similar to Reference Implementation | Pretrained agent available
---------|---------|---------|---------
| **RoboSchool Envs** |
InvertedPendulumPyBulletEnv-v0 | Yes | Yes | No
InvertedDoublePendulumPyBulletEnv-v0 | Yes | Yes | No
InvertedPendulumSwingupPyBulletEnv-v0 | Yes | Yes | No
ReacherPyBulletEnv-v0 | Yes | Yes | No
Walker2DPyBulletEnv-v0 | Yes | No | No
HalfCheetahPyBulletEnv-v0 | Yes | No | No
AntPyBulletEnv-v0 | Yes | Yes | No
HopperPyBulletEnv-v0 | Yes | Yes | No
HumanoidPyBulletEnv-v0 | Yes | Yes | No
HumanoidFlagrunPyBulletEnv-v0 | Yes | Yes | No
HumanoidFlagrunHarderPyBulletEnv-v0 | Yes | Yes | No
AtlasPyBulletEnv-v0 | WIP | No | No
PusherPyBulletEnv-v0 | WIP | No | No
ThrowerPyBulletEnv-v0 | WIP | No | No
StrikerPyBulletEnv-v0 | WIP | No | No
| **MuJoCo Envs** |
InvertedPendulumMuJoCoEnv-v0 | Yes | Yes | Yes
InvertedDoublePendulumMuJoCoEnv-v0 | Yes | Yes | Yes
ReacherMuJoCoEnv-v0 | No | No | No
Walker2DMuJoCoEnv-v0 | Yes | No | No
HalfCheetahMuJoCoEnv-v0 | Yes | No | No
AntMuJoCoEnv-v0 | Yes | No | No
HopperMuJoCoEnv-v0 | Yes | No | No
HumanoidMuJoCoEnv-v0 | Yes | No | No
PusherMuJoCoEnv-v0 | No | No | No
ThrowerMuJoCoEnv-v0 | No | No | No
StrikerMuJoCoEnv-v0 | No | No | No
[See What's New section below](#What's New)
## Basics
(taken from OpenAI gym readme)
There are two basic concepts in reinforcement learning: the
environment (namely, the outside world) and the agent (namely, the
algorithm you are writing). The agent sends `actions` to the
environment, and the environment repliExcerpt of 8,252 characters
Read on GitHub
Benjamin Ellenberger · IDSC.io · Switzerland
41
Gabriel Huang · @ServiceNow · Canada
3
1
1
1
1
1
1
1
Lucas Vazquez
1
1
Silvio Traversaro · Generative Bionics @gbionics
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:bc9a60467f23effe, topic:openai-gym, desc:reinforcement learning, readme:reinforcement learning