search

benelot / pybullet-gym

Open-source implementations of OpenAI Gym MuJoCo environments for use with the OpenAI Gym Reinforcement Learning Research Platform.
https://pybullet.org/
Other
819 stars 124 forks source link

v2 environments from openai gym #12

Open novinsh opened 5 years ago

novinsh commented 5 years ago

I need some other environments from the openai gym. Any guidelines to help to port the rest of the environments from there?

benelot commented 5 years ago

Which ones do you need? It is basically as simple/hard as the following: Copy the roboschool code and change the number of observations/actions to the number in the mujoco implementations of OpenAI. Then, and this is the most problematic part, find the observations from bullet which correspond to the ones in the mujoco implementations. Finally, to ensure compatibility, check if an agent trained in mujoco behaves similarly on your implementation. The last step could theoretically be a separate step, I would be fine if you found the corresponding observation signals.

On Sat, Sep 15, 2018, 13:30 Novin Shahroudi notifications@github.com wrote:

I need some other environments from the openai gym. Any guidelines to help to port the rest of the environments from there?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/benelot/pybullet-gym/issues/12, or mute the thread https://github.com/notifications/unsubscribe-auth/AC97q3ZX9CIGeeOlvq8FmLmK336HbDocks5ubOTHgaJpZM4WqUXN .

benelot commented 5 years ago

I just finished the first step i described above. The envs work, but contain a lot of dummy observations, because I do not know the corresponding observations in bullet physics.