search

semitable / robotic-warehouse

Multi-Robot Warehouse (RWARE): A multi-agent reinforcement learning environment
MIT License
292 stars 69 forks source link

Questions of reward every step output #14

Open POJIANWU opened 2 years ago

POJIANWU commented 2 years ago

Hi,I'm interested in your project,but I'm confused about your warehouse.py,that is,why does't it output reward values every time step?Longing for your reply!

semitable commented 2 years ago

It should give you rewards at every time step (at a typical gym-interface fashion).

Do note that this is a (somewhat) sparse-reward environment. Non-zero rewards are given only upon successfully delivering a requested shelf.

POJIANWU commented 1 year ago

It should give you rewards at every time step (at a typical gym-interface fashion).

Do note that this is a (somewhat) sparse-reward environment. Non-zero rewards are given only upon successfully delivering a requested shelf.

Thanks for your reply!Another question that is when I run the test_obs_space_2 in test_env.py,I meet this error"assert len(gym.spaces.flatten(s, o)) == env._obs_length E assert 133 == 116".What should I do?Hope for your soon reply!