rofunc.learning.RofuncRL.utils.elegantrl_utils

Contents

`rofunc.learning.RofuncRL.utils.elegantrl_utils`#

1. Module Contents#

1.1. Classes#

ElegantRLIsaacGymEnvWrapper

1.2. Functions#

1.3. API#

class rofunc.learning.RofuncRL.utils.elegantrl_utils.ElegantRLIsaacGymEnvWrapper(env, cfg, should_print=False)#

Initialization

reset() → torch.Tensor#

Resets the environments in the VecTask that need to be reset.

Returns:: torch.Tensor: the next states in the simulation.

step(actions: torch.Tensor) → Tuple[torch.Tensor, torch.Tensor, torch.Tensor, Dict]#

Steps through the vectorized environment.

Args:

actions (torch.Tensor): a multidimensional tensor of actions to perform on: each environment.

Returns:

Tuple[torch.Tensor, torch.Tensor, torch.Tensor, Dict]: a tuple containing: observations, rewards, dones, and extra info.

rofunc.learning.RofuncRL.utils.elegantrl_utils.setup(custom_args, eval_mode=False)#