Apply suggestions from code review

src/primaite/agents/agent.py

@@ -50,7 +50,7 @@ class AgentSessionABC(ABC):
     def __init__(self, training_config_path, lay_down_config_path):
         if not isinstance(training_config_path, Path):
             training_config_path = Path(training_config_path)
-        self._training_config_path: Final[Union[Path]] = training_config_path
+        self._training_config_path: Final[Union[Path, str]] = training_config_path
         self._training_config: Final[TrainingConfig] = training_config.load(self._training_config_path)
 
         if not isinstance(lay_down_config_path, Path):

src/primaite/agents/rllib.py

@@ -55,7 +55,7 @@ class RLlibAgent(AgentSessionABC):
             msg = "Expected PPO or A2C agent_identifier, " f"got {self._training_config.agent_identifier.value}"
             _LOGGER.error(msg)
             raise ValueError(msg)
-        self._agent_config: PPOConfig
+        self._agent_config: Union[PPOConfig, A2CConfig]
 
         self._current_result: dict
         self._setup()

src/primaite/environment/observations.py

@@ -57,7 +57,7 @@ class NodeLinkTable(AbstractObservationComponent):
     """
 
     _FIXED_PARAMETERS: int = 4
-    _MAX_VAL: int = 1_000_000
+    _MAX_VAL: int = 1_000_000_000
     _DATA_TYPE: type = np.int64
 
     def __init__(self, env: "Primaite"):