huangwl18/modular-rl
[ICML 2020] PyTorch Code for "One Policy to Control Them All: Shared Modular Policies for Agent-Agnostic Control"
This project helps robotics researchers and engineers train a single, versatile control policy that can manage a wide variety of simulated robotic agents with different physical forms. Instead of training a new policy for each robot, you input diverse robot morphologies (as MuJoCo XML files), and the output is one policy capable of controlling all of them, even new, unseen designs. This is for professionals working on developing generalizable control systems for robotics.
232 stars. No commits in the last 6 months.
Use this if you need to train a single reinforcement learning policy to control many different simulated robotic agents, especially when the exact robot morphology might change or be unknown beforehand.
Not ideal if your application involves physical robots or non-2D planar agents, as the current implementation is focused on simulated 2D planar MuJoCo environments.
Stars
232
Forks
34
Language
Jupyter Notebook
License
—
Category
Last pushed
Dec 27, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/huangwl18/modular-rl"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
DLR-RM/stable-baselines3
PyTorch version of Stable Baselines, reliable implementations of reinforcement learning algorithms.
google-deepmind/dm_control
Google DeepMind's software stack for physics-based simulation and Reinforcement Learning...
Denys88/rl_games
RL implementations
pytorch/rl
A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.
yandexdataschool/Practical_RL
A course in reinforcement learning in the wild