#917 - Synced with dev and added better logging

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

@@ -38,26 +38,23 @@ hard_coded_agent_view: FULL
 action_type: ANY
 
 # Number of episodes to run per session
-num_episodes: 100
+num_episodes: 10
 
 # Number of time_steps per episode
 num_steps: 256
 
 # Sets how often the agent will save a checkpoint (every n time episodes).
 # Set to 0 if no checkpoints are required. Default is 10
-checkpoint_every_n_episodes: 100
+checkpoint_every_n_episodes: 10
 
 # Time delay between steps (for generic agents)
-time_delay: 3
+time_delay: 5
 
-# 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]
+# Type of session to be run. Options are:
+# "TRAIN" (Trains an agent)
+# "EVAL" (Evaluates an agent)
+# "TRAIN_EVAL" (Trains then evaluates an agent)
+session_type: TRAIN
 
 # Environment config values
 # The high value for the observation space

src/primaite/config/training_config.py

@@ -69,7 +69,7 @@ class TrainingConfig:
     "The delay between steps (ms). Applies to generic agents only"
 
     # file
-    session_type: SessionType = SessionType.TRAINING
+    session_type: SessionType = SessionType.TRAIN
     "The type of PrimAITE session to run"
 
     load_agent: str = False
@@ -171,6 +171,7 @@ class TrainingConfig:
     file_system_scanning_limit: int = 5
     "The time taken to scan the file system"
 
+
     @classmethod
     def from_dict(
             cls,
@@ -183,7 +184,7 @@ class TrainingConfig:
             "action_type": ActionType,
             "session_type": SessionType,
             "output_verbose_level": OutputVerboseLevel,
-            "hard_coded_agent_view": HardCodedAgentView
+            "hard_coded_agent_view": HardCodedAgentView,
         }
 
         for field, enum_class in field_enum_map.items():
@@ -211,6 +212,20 @@ class TrainingConfig:
 
         return data
 
+    def __str__(self) -> str:
+        tc = f"TrainingConfig(agent_framework={self.agent_framework.name}, "
+        if self.agent_framework is AgentFramework.RLLIB:
+            tc += f"deep_learning_framework=" \
+                  f"{self.deep_learning_framework.name}, "
+        tc += f"agent_identifier={self.agent_identifier.name}, "
+        if self.agent_identifier is AgentIdentifier.HARDCODED:
+            tc += f"hard_coded_agent_view={self.hard_coded_agent_view.name}, "
+        tc += f"action_type={self.action_type.name}, "
+        tc += f"observation_space={self.observation_space}, "
+        tc += f"num_episodes={self.num_episodes}, "
+        tc += f"num_steps={self.num_steps})"
+        return tc
+
 
 def load(
         file_path: Union[str, Path],