dm_control suite for OpenAI Gym env
Project description
Introduction
I simplified martinseilair's dm_control2gym
as well as made it compatible with the latest MuJoCo
library(e.g., 2.0.0). .... yeah,, I will upload the documents when I have time.
Usage
-
MDP tasks:
env
returns a state of a robot at each time-stepimport itertools from dm_control2gym.util import make_dm2gym_env_state env = make_dm2gym_env_state(env_name="cheetah_run") state = env.reset() print("State shape: ", state.shape) total_reward = 0 for t in itertools.count(): action = env.action_space.sample() state, reward, done, _ = env.step(action) total_reward += reward if done: break env.close() print("Total Reward: {}".format(total_reward))
-
POMDP tasks:
env
returns a raw image observation at each time-stepimport itertools from dm_control2gym.util import make_dm2gym_env_obs env = make_dm2gym_env_obs(env_name="cheetah_run", num_repeat_action=1) obs = env.reset() print("Obs shape: ", obs.shape) total_reward = 0 for t in itertools.count(): action = env.action_space.sample() obs, reward, done, _ = env.step(action) total_reward += reward if done: break env.close() print("Total Reward: {}".format(total_reward))
For more details
Please refer to martinseilair's dm_control2gym
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
dm_control2gym-0.0.1.tar.gz
(5.1 kB
view hashes)
Built Distribution
Close
Hashes for dm_control2gym-0.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 49304fa62f533ff75a3c2ee20b2ae60c296a9b42e466b7dc5e619f59195f86bf |
|
MD5 | 1fe26e6c5a0a1588a95acd33657ca05c |
|
BLAKE2b-256 | febf7481eef2bea3116679629181fc7df3113c10a7f2b457b02970f777603b79 |