peg-in-hole-gym

Gym env for multi-task simulation with kinds of robotic arms, engined by pybullet

Task List

• peg-in-hole

  

• More

  With meta_env you can easily create your custom env.

Features

• multi-agent in one simulation
• multi-simulation with multiprocessing
• alternative for kinds of robotic arms (panda and ur5)

Installation

$ git clone [email protected]:guodashun/peg-in-hole-gym.git
$ cd peg-in-hole-gym
$ pip install -e .

Usage

Initial the environment(e.g. random-fly)

import gym
import pybullet as p
import peg_in_hole_gym

env = gym.make('peg-in-hole-mp-v0', client=p.GUI, task='random-fly', mp_num=mp_num, sub_num=sub_num, offset = [2.,3.,0.],args=['Banana', 1/120.], is_test=True)
env.reset()
env.render()

Perform a task

obs = env.reset()
while True:
    obs, reward, done, info = env.step(env.action_space.sample())
	time.sleep(0.01)

Reference

• PyBullet Quickstart Guide