Giter Club home page Giter Club logo

rainbow_ddpg's Introduction

Rainbow DDPG

This repository contains Rainbow DDPG algorithm from paper Sim-to-Real Reinforcement Learning for Deformable Object Manipulation along with a toy pushing task to demonstrate how to use the code.

Instructions

The code was tested on Mac OS with Python3.6. Use of virtualenvs is recommended. To run:

pip install -r requirements.txt
python main.py

Runnign a full training may take more than 24 hours on a machine with Nvidia Titan GPU and use a considerable amount of memory.

To run a demonstration of the toy task:

pip install -r requirements.txt
python run_demo.py

Please note that the hyper parameters are not necessarily optimised for the task.

References

For a complete list of references, please see the accompanying paper.

The learning algorithm is based on OpenAI baselines (https://github.com/openai/baselines), the perlin noise file is heavily based on https://github.com/nikagra/python-noise/blob/master/noise.py and robot meshes are generated from https://github.com/Kinovarobotics/kinova-ros.

rainbow_ddpg's People

Contributors

janmatas avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar

rainbow_ddpg's Issues

Process killed?

Hi, I'm trying to run your project (more specifically, 'main.py' ) and everything goes smoothly, however, after 17 epoches, the process is 'killed'. I'm running this program in a GTX 1050Ti, is my GPU too lame to running this project?

Gym environment "Pusher-v1"

Hi Matas,

I am trying to reproduce your work on ubuntu 18.04. I found it is extremely interesting.

But I encountered the following problem. After I debug the demo, I found that

env = gym.make('Pusher-v1')

does not work anymore. Gym seems to only provide 'Pusher-v2'.

Even I installed gym==0.10.5, gym did not seem to find 'Pusher-v1'.

Is there other methods to workaround to reproduce your work.

program stuck at calculateInverseKinematics

Hi, I was trying to run your code but it always stuck during env.step(action), for both main.py and run_demo.py. I identified the root cause is that it took forever for function calculateInverseKinematics to finish in Mico.py (in function compute_ik_poses). Could you help me with this? Thanks!

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.