https://zhuanlan.zhihu.com/p/33834336
Environment for Agents
OpenAI GYM provides an environment for testing algorithms, and it interfaces with many famous test problems like
- Physical control
- Atari
- Mojuco
https://github.com/openai/gym/issues/1726#issuecomment-550580367
https://zhuanlan.zhihu.com/p/33834336
https://github.com/Kojoley/atari-py
pip install --no-index -f https://github.com/Kojoley/atari-py/releases atari_py
pip install gym
MoJuCo
https://web.stanford.edu/class/cs234/assignment3/mujoco_win_install.pdf
Note seems older versions of gym + older version of MuJoCo + older version of mujoco_py will be compatible. For example
mujoco๏ผ131 mujoco_py:0.5.7 tensorflow:1.5.0 gym:0.9.1 python3.6
but you may need to use the order version of environment models in mujoco platform. (eg. you have to use ‘InvertedPendulum-v1’ in stead of the higher version’InvertedPendulum-v2’) Good luck~
You may need to resolve some error using it.
pip install gym==0.9.1 --use-feature=2020-resolver
https://github.com/openai/mujoco-py/issues/261
Test Rendering
https://www.endtoend.ai/envs/gym/mujoco/
import gym
env = gym.make('Humanoid-v2')
env.reset()
env.render()
observation = env.reset()
for t in range(100):
env.render()
print(observation)
action = env.action_space.sample()
observation, reward, done, info = env.step(action)
if done:
print("Episode finished after {} timesteps".format(t+1))
break
Build Models in Mujoco
https://studywolf.wordpress.com/2020/03/22/building-models-in-mujoco/
