rosewang2008 / gym-cooking Goto Github PK

Greetings.

I was quite inspired by your work on theory of mind and it;s integration with AI agents.
Thank you very much for publishing the experiments, as well as the experiment simulator.

Regarding the latter, I was hoping you could provide a bit more details regarding the action space for the overcooked environment featured in this repository.

I went as far as looking into the OvercookedEnvironment class's step method, but it is not clear what I should feed when calling env.step(action_dict).

The paper mentions an action space size 5 to move the agent: North, East, West, South, Stay still.
I tried passing action_dict={"agent-1": 0,1,2,3,4... etc...} but it did not seem to work consistently.

For a bit more context, I am trying to use this environment with off-the-shelf RL algorithms such as those provided in Stable Baselines 3, so it would be really helpful to have a specification for env.action_space.

Looking forward to hear from you.
Best regards.

Congratulations on your nice paper and the award! I noticed a small typo in the folder gym_cooking/misc/game/graphics - plate.png should be renamed to Plate.png

Hello,

I've forked and clone your repo and tried to run the command python main.py --num-agents 2 --level partial-divider_salad --model1 bd --model2 bd but am not able to run it successfully:

Will it be possible for you to help?

Thank you.

Hi,
This looks great, I can't wait to try it!
However, when I tried to run as described in the documentation, I get an error about a missing module misc.

(gym) $ python main.py --num-agents 4 --level partial-divider_salad --model1 greedy --model2 bd --model3 fb --model4 dc
Traceback (most recent call last):
  File "main.py", line 7, in <module>
    from misc.game.gameplay import GamePlay
ModuleNotFoundError: No module named 'misc'

am I missing something? or doing something wrong here?
thanks!

Hi Rose,

Great project! I really enjoyed the images visualizing each step.

However, I ran into the following error when trying to run experiments with the --with-image-obs and --record commands:

I believe the issue lies in the following part of the overcooked_environment.py file, where the new_obs variable is used for two different objects - the OvercookedEnvironment and a GameImage:

When I renamed the new_obs variables on lines 210 and 213, everything worked fine.

README suggests a run_experiments.sh file, looks like it's sitting in your .gitignore and thus not pushed to GitHub.

IIUC, the environment support only one recipe for one layout.
So my goal is to have multiple recipes for one layout in combination with order queue that prioritize earlier order with higher reward.

Is there a specific file that support this feature? If not is there a recommended way to implement this?

'OvercookedEnvironment' object has no attribute 'game'

I run the code with the following command as instructed:
python main.py --num-agents 3 --level full-divider_tl --model1 up --model2 dc --model3 bd

It seems self.game is only created with record option or with_image_obs option