rofunc.learning.RofuncRL.agents.offline.dtrans_agent#

1.  Module Contents#

1.1.  Classes#

DTransAgent

Decision Transformer (DTrans) Agent

1.2.  API#

class rofunc.learning.RofuncRL.agents.offline.dtrans_agent.DTransAgent(cfg: omegaconf.DictConfig, observation_space: Optional[Union[int, Tuple[int], gym.Space, gymnasium.Space]], action_space: Optional[Union[int, Tuple[int], gym.Space, gymnasium.Space]], device: Optional[Union[str, torch.device]] = None, experiment_dir: Optional[str] = None, rofunc_logger: Optional[rofunc.logger.BeautyLogger] = None)#

Bases: rofunc.learning.RofuncRL.agents.base_agent.BaseAgent

Decision Transformer (DTrans) Agent

“Decision Transformer: Reinforcement Learning via Sequence Modeling”. Chen et al. 2021. https://arxiv.org/abs/2106.01345

Rofunc documentation: https://rofunc.readthedocs.io/en/latest/lfd/RofuncRL/DTrans.html

Initialization

Parameters:

  • cfg – Configurations

  • observation_space – Observation space

  • action_space – Action space

  • device – Device on which the torch tensor is allocated

  • experiment_dir – Directory for storing experiment data

  • rofunc_logger – Rofunc logger

checkpoint_best_modules = None#

Get hyper-parameters from config

act(states, actions, rewards, returns_to_go, timesteps)#
update_net(batch)#