releso.agent.PretrainedAgent

class releso.agent.PretrainedAgent(*, save_location: Path, logger_name: str | None = None, tensorboard_log: str | None = None, type: Literal['PPO', 'SAC', 'DDPG', 'A2C', 'DQN'], path: FilePath | Path, tesorboard_run_directory: str | Path | None = None)

Bases: BaseAgent

Pretrained agent definition.

This class can be used to load pretrained agents, instead of using untrained agents. Can also be used to only validate this agent without training it further. Please see validation section for this use-case.

__init__(**data: Any) → None: Constructor for the ReLeSO basemodel object.

Methods

`convert_to_pathlib_add_datetime`(v)	Add timestamp to save_location, of applicable.
`get_agent`(environment[, normalizer_divisor])	Tries to locate the agent defined and to load it correctly.
`get_logger`()	Gets the currently defined environment logger.
`get_next_tensorboard_experiment_name`()	Return the name of the tensorboard experiment.
`set_logger_name_recursively`(logger_name)	Set the logger_name variable for all child elements.

Attributes

`agent_type`	What RL algorithm was used to train the agent.
`path`	Path to the save files of the pretrained agent.
`tesorboard_run_directory`
`tensorboard_log`	base directory of the tensorboard logs if given an experiment name with a current timestamp is also added.
`save_location`	Definition of the save location of the logs and validation results.
`logger_name`	name of the logger.

agent_type: Literal['PPO', 'SAC', 'DDPG', 'A2C', 'DQN']: What RL algorithm was used to train the agent. Needs to be know to correctly load the agent.

get_agent(environment: GymEnvironment, normalizer_divisor: int = 1) → BaseAlgorithm

Tries to locate the agent defined and to load it correctly.

Parameters:

environment (GymEnvironment) – Environment with which the agent will
interact.
normalizer_divisor (int) – Currently not used in this function.

Raises:

AgentUnknownException – Possible exception which can be
thrown. –

Returns:

Return the correctly loaded agent.

Return type:

BaseAlgorithm

get_next_tensorboard_experiment_name() → str | None

Return the name of the tensorboard experiment.

The tensorboard experiment name of the original training run if given else a new one with the current time stamp.

Returns:: tensorboard experiment name
Return type:: str

path: FilePath | Path: Path to the save files of the pretrained agent.