Table of Contents
Shortcuts

openrl.selfplay.opponents package

Submodules

openrl.selfplay.opponents.base_opponent module

class openrl.selfplay.opponents.base_opponent.BaseOpponent(opponent_id: Optional[str] = None, opponent_path: Optional[Union[str, pathlib.Path]] = None, opponent_info: Optional[Dict[str, str]] = None)[source]

Bases: abc.ABC

abstract act(player_name, observation, reward, termination, truncation, info)[source]
load(opponent_path: Union[str, pathlib.Path], opponent_id: Optional[str] = None) → openrl.selfplay.opponents.base_opponent.BaseOpponent[source]
log()[source]
property opponent_type
reset(env=None, opponent_player: Optional[str] = None)[source]
set_env(env, opponent_player: Optional[str] = None)[source]

openrl.selfplay.opponents.jidi_opponent module

class openrl.selfplay.opponents.jidi_opponent.JiDiOpponent(opponent_id: Optional[str] = None, opponent_path: Optional[Union[str, pathlib.Path]] = None, opponent_info: Optional[Dict[str, str]] = None, jidi_controller: Optional[Callable] = None, player_num: int = 1)[source]

Bases: openrl.selfplay.opponents.base_opponent.BaseOpponent

act(player_name, observation, reward, termination, truncation, info)[source]

openrl.selfplay.opponents.network_opponent module

class openrl.selfplay.opponents.network_opponent.NetworkOpponent(opponent_id: str, opponent_path: Union[str, pathlib.Path], opponent_info: Dict[str, str])[source]

Bases: openrl.selfplay.opponents.base_opponent.BaseOpponent

act(player_name, observation, reward, termination, truncation, info)[source]
reset(env=None, opponent_player: Optional[str] = None)[source]

openrl.selfplay.opponents.opponent_env module

class openrl.selfplay.opponents.opponent_env.BaseOpponentEnv(env, opponent_player: str)[source]

Bases: object

property action_space: Union[gymnasium.spaces.space.Space[gymnasium.core.ActType], gymnasium.spaces.space.Space[gymnasium.core.WrapperActType]]
property agent_num: int
property observation_space: Union[gymnasium.spaces.space.Space[gymnasium.core.ObsType], gymnasium.spaces.space.Space[gymnasium.core.WrapperObsType]]
property parallel_env_num: int
process_action(action)[source]
process_obs(observation, termination, truncation, info)[source]
reset(**kwargs)[source]

openrl.selfplay.opponents.opponent_template module

class openrl.selfplay.opponents.opponent_template.OpponentTemplate(opponent_template: Union[str, pathlib.Path], copy_script_file: bool = False)[source]

Bases: object

check()[source]
load_opponent_info() → Dict[str, Any][source]
load_script_files() → Union[List[str], List[pathlib.Path]][source]
save(opponent_path: Union[str, pathlib.Path], opponent_info: Dict[str, Any])[source]

openrl.selfplay.opponents.random_opponent module

class openrl.selfplay.opponents.random_opponent.RandomOpponent(opponent_id: Optional[str] = None, opponent_path: Optional[Union[str, pathlib.Path]] = None, opponent_info: Optional[Dict[str, str]] = None)[source]

Bases: openrl.selfplay.opponents.base_opponent.BaseOpponent

act(player_name, observation, reward, termination, truncation, info)[source]
sample_random_action(player_name, observation, reward, termination, truncation, info)[source]

openrl.selfplay.opponents.utils module

openrl.selfplay.opponents.utils.check_opponent_template(opponent_template: Union[str, pathlib.Path])[source]
openrl.selfplay.opponents.utils.get_opponent_from_info(opponent_info: Dict[str, str], current_opponent: Optional[openrl.selfplay.opponents.base_opponent.BaseOpponent] = None, lazy_load_opponent: bool = True) → Tuple[Optional[openrl.selfplay.opponents.base_opponent.BaseOpponent], bool][source]
openrl.selfplay.opponents.utils.get_opponent_from_path(opponent_path: Union[str, pathlib.Path], current_opponent: Optional[openrl.selfplay.opponents.base_opponent.BaseOpponent] = None, lazy_load_opponent: bool = True, opponent_info: Optional[Dict[str, str]] = None) → Tuple[Optional[openrl.selfplay.opponents.base_opponent.BaseOpponent], bool][source]
openrl.selfplay.opponents.utils.get_opponent_id(opponent_info: Optional[Dict[str, str]]) → Optional[str][source]
openrl.selfplay.opponents.utils.get_opponent_info(info_path: Optional[Union[str, pathlib.Path]]) → Optional[Dict[str, str]][source]
openrl.selfplay.opponents.utils.load_opponent_from_jidi_path(opponent_path: Union[str, pathlib.Path], opponent_info: Optional[Dict[str, str]] = None, player_num: int = 1) → Optional[openrl.selfplay.opponents.base_opponent.BaseOpponent][source]
openrl.selfplay.opponents.utils.load_opponent_from_path(opponent_path: Union[str, pathlib.Path], opponent_info: Optional[Dict[str, str]] = None) → Optional[openrl.selfplay.opponents.base_opponent.BaseOpponent][source]

Module contents