openrl.drivers package

Submodules

openrl.drivers.base_driver module

class openrl.drivers.base_driver.BaseDriver

Bases: abc.ABC

abstract run(*args, **kwargs)

openrl.drivers.offpolicy_driver module

class openrl.drivers.offpolicy_driver.OffPolicyDriver(config: Dict[str, Any], trainer, buffer, agent, rank: int = 0, world_size: int = 1, client=None, logger: Optional[openrl.utils.logger.Logger] = None)

Bases: openrl.drivers.rl_driver.RLDriver

act(step: int)

actor_rollout()

add2buffer(data)

compute_returns()

openrl.drivers.onpolicy_driver module

class openrl.drivers.onpolicy_driver.OnPolicyDriver(config: Dict[str, Any], trainer, buffer, agent, rank: int = 0, world_size: int = 1, client=None, logger: Optional[openrl.utils.logger.Logger] = None, callback: Union[None, Callable, List[openrl.utils.callbacks.callbacks.BaseCallback], openrl.utils.callbacks.callbacks.BaseCallback] = None)

Bases: openrl.drivers.rl_driver.RLDriver

act(step: int)

actor_rollout() → Tuple[Dict[str, Any], bool]

add2buffer(data)

compute_returns()

openrl.drivers.rl_driver module

class openrl.drivers.rl_driver.RLDriver(config: Dict[str, Any], trainer, buffer, agent, rank: int = 0, world_size: int = 1, client=None, logger: Optional[openrl.utils.logger.Logger] = None, callback: Union[None, Callable, List[openrl.utils.callbacks.callbacks.BaseCallback], openrl.utils.callbacks.callbacks.BaseCallback] = None)

Bases: openrl.drivers.base_driver.BaseDriver, abc.ABC

abstract act(step: int)

abstract actor_rollout()

abstract add2buffer(data)

abstract compute_returns()

learner_update()

reset_and_buffer_init()

run() → None

Module contents