#1386: remove unneeded configs + setting the seed globally + temp test

src/primaite/config/_package_data/training/training_config_main.yaml

@@ -31,6 +31,16 @@ agent_identifier: PPO
 # False
 random_red_agent: False
 
+# The (integer) seed to be used in random number generation
+# Default is None (null)
+seed: null
+
+# Set whether the agent will be deterministic instead of stochastic
+# Options are:
+# True
+# False
+deterministic: False
+
 # Sets what view of the environment the deterministic hardcoded agent has. The default is BASIC.
 # Options are:
 # "BASIC" (The current observation space only)

src/primaite/config/_package_data/training/training_config_random_red_agent.yaml

@@ -1,101 +0,0 @@
-# Main Config File
-
-# Generic config values
-# Choose one of these (dependent on Agent being trained)
-# "STABLE_BASELINES3_PPO"
-# "STABLE_BASELINES3_A2C"
-# "GENERIC"
-agent_identifier: STABLE_BASELINES3_A2C
-
-# Sets whether Red Agent POL and IER is randomised.
-# Options are:
-# True
-# False
-random_red_agent: True
-
-# Sets How the Action Space is defined:
-# "NODE"
-# "ACL"
-# "ANY" node and acl actions
-action_type: NODE
-# Number of episodes to run per session
-num_episodes: 10
-# Number of time_steps per episode
-num_steps: 256
-# Time delay between steps (for generic agents)
-time_delay: 10
-# Type of session to be run (TRAINING or EVALUATION)
-session_type: TRAINING
-# Determine whether to load an agent from file
-load_agent: False
-# File path and file name of agent if you're loading one in
-agent_load_file: C:\[Path]\[agent_saved_filename.zip]
-
-# Environment config values
-# The high value for the observation space
-observation_space_high_value: 1000000000
-
-# Reward values
-# Generic
-all_ok: 0
-# Node Hardware State
-off_should_be_on: -10
-off_should_be_resetting: -5
-on_should_be_off: -2
-on_should_be_resetting: -5
-resetting_should_be_on: -5
-resetting_should_be_off: -2
-resetting: -3
-# Node Software or Service State
-good_should_be_patching: 2
-good_should_be_compromised: 5
-good_should_be_overwhelmed: 5
-patching_should_be_good: -5
-patching_should_be_compromised: 2
-patching_should_be_overwhelmed: 2
-patching: -3
-compromised_should_be_good: -20
-compromised_should_be_patching: -20
-compromised_should_be_overwhelmed: -20
-compromised: -20
-overwhelmed_should_be_good: -20
-overwhelmed_should_be_patching: -20
-overwhelmed_should_be_compromised: -20
-overwhelmed: -20
-# Node File System State
-good_should_be_repairing: 2
-good_should_be_restoring: 2
-good_should_be_corrupt: 5
-good_should_be_destroyed: 10
-repairing_should_be_good: -5
-repairing_should_be_restoring: 2
-repairing_should_be_corrupt: 2
-repairing_should_be_destroyed: 0
-repairing: -3
-restoring_should_be_good: -10
-restoring_should_be_repairing: -2
-restoring_should_be_corrupt: 1
-restoring_should_be_destroyed: 2
-restoring: -6
-corrupt_should_be_good: -10
-corrupt_should_be_repairing: -10
-corrupt_should_be_restoring: -10
-corrupt_should_be_destroyed: 2
-corrupt: -10
-destroyed_should_be_good: -20
-destroyed_should_be_repairing: -20
-destroyed_should_be_restoring: -20
-destroyed_should_be_corrupt: -20
-destroyed: -20
-scanning: -2
-# IER status
-red_ier_running: -5
-green_ier_blocked: -10
-
-# Patching / Reset durations
-os_patching_duration: 5            # The time taken to patch the OS
-node_reset_duration: 5             # The time taken to reset a node (hardware)
-service_patching_duration: 5       # The time taken to patch a service
-file_system_repairing_limit: 5      # The time take to repair the file system
-file_system_restoring_limit: 5      # The time take to restore the file system
-file_system_scanning_limit: 5       # The time taken to scan the file system

src/primaite/config/training_config.py

@@ -178,6 +178,12 @@ class TrainingConfig:
     file_system_scanning_limit: int = 5
     "The time taken to scan the file system"
 
+    deterministic: bool = False
+    "If true, the training will be deterministic"
+
+    seed: Optional[int] = None
+    "The random number generator seed to be used while training the agent"
+
     @classmethod
     def from_dict(cls, config_dict: Dict[str, Union[str, int, bool]]) -> TrainingConfig:
         """
@@ -201,12 +207,6 @@ class TrainingConfig:
                 config_dict[key] = value[config_dict[key]]
         return TrainingConfig(**config_dict)
 
-    deterministic: bool = False
-    "If true, the training will be deterministic"
-
-    seed: Optional[int] = None
-    "The random number generator seed to be used while training the agent"
-
     def to_dict(self, json_serializable: bool = True):
         """
         Serialise the ``TrainingConfig`` as dict.

src/primaite/environment/primaite_env.py

@@ -2,6 +2,7 @@
 """Main environment module containing the PRIMmary AI Training Evironment (Primaite) class."""
 import copy
 import logging
+import random
 import uuid as uuid
 from pathlib import Path
 from random import choice, randint, sample, uniform
@@ -218,7 +219,7 @@ class Primaite(Env):
             # [0, 3] - action on property (0 = nothing, On / Scan, Off / Repair, Reset / Patch / Restore) # noqa
             # [0, num services] - resolves to service ID (0 = nothing, resolves to service) # noqa
             self.action_dict = self.create_node_action_dict()
-            self.action_space = spaces.Discrete(len(self.action_dict), seed=self.training_config.seed)
+            self.action_space = spaces.Discrete(len(self.action_dict))
         elif self.training_config.action_type == ActionType.ACL:
             _LOGGER.debug("Action space type ACL selected")
             # Terms (for ACL action space):
@@ -229,17 +230,22 @@ class Primaite(Env):
             # [0, num services] - Protocol (0 = any, then 1 -> x resolving to protocol)
             # [0, num ports] - Port (0 = any, then 1 -> x resolving to port)
             self.action_dict = self.create_acl_action_dict()
-            self.action_space = spaces.Discrete(len(self.action_dict), seed=self.training_config.seed)
+            self.action_space = spaces.Discrete(len(self.action_dict))
         elif self.training_config.action_type == ActionType.ANY:
             _LOGGER.debug("Action space type ANY selected - Node + ACL")
             self.action_dict = self.create_node_and_acl_action_dict()
-            self.action_space = spaces.Discrete(len(self.action_dict), seed=self.training_config.seed)
+            self.action_space = spaces.Discrete(len(self.action_dict))
         else:
             _LOGGER.error(f"Invalid action type selected: {self.training_config.action_type}")
 
         self.episode_av_reward_writer = SessionOutputWriter(self, transaction_writer=False, learning_session=True)
         self.transaction_writer = SessionOutputWriter(self, transaction_writer=True, learning_session=True)
 
+        # set the seed globally if there is one
+        if self.training_config.seed:
+            random.seed(self.training_config.seed)
+            np.random.seed(self.training_config.seed)
+
     @property
     def actual_episode_count(self) -> int:
         """Shifts the episode_count by -1 for RLlib."""